Think of it the way the U.S. dollar works–every dollar is equivalent to and interchangeable with any other dollar. ERC–20 enables similar interchangeability for digital tokens. This matters because it gives token contracts a common interface that wallets, exchanges, and apps can support consistently and seamlessly. In this post we dive into how ERC–20 tokens work and the ever growing types of use cases they can support.

Let’s Get Started!

Originally proposed on November 19, 2015, by developers Fabian Vogelsteller and Vitalik Buterin, (the co–founder of Ethereum) the ERC–20 token standard introduced a common interface for fungible tokens within smart contracts.

ERC–20 is one of the most established token standards in crypto and is the default choice for many Ethereum–based assets. It is a widely used standard for token applications that work with other products and services, and many major stablecoins such as USDC, UNI, and AAVE use it. Adoption is especially strong for assets that need broad wallet and exchange support, including stablecoins and DeFi governance tokens. The standard’s compatibility with the wider Ethereum ecosystem has made it a foundational building block for token issuance and liquidity on the Ethereum block chain.

Before ERC–20, developers had to create unique codebases and custom logic for every new token they wanted to deploy on the Ethereum blockchain. This lack of standardization led to a highly fragmented, incompatible and hence very difficult to scale landscape. ERC–20 addressed this problem by providing a unified set of rules that simplifies interaction between tokens and applications.

ERC–20 was formally codified into Ethereum Improvement Proposal 20 (EIP–20) in 2017.

ERC–20 Tokens: How They Work–Let’s Go Under the Hood.

An ERC–20 token is implemented as a smart contract in software. It is a self–executing software script that lives on the Ethereum blockchain ledger. The smart contract maintains a database of Ethereum addresses and their corresponding balances. When you “send” 10 tokens, you are sending a message to the smart contract saying: “Please subtract 10 tokens from my address and add 10 tokens to this other address.” This simple elegance is a big part of why the standard has prospered.

ERC–20 specifies a set of functions and events that must be implemented into a token so wallets, exchanges, and other contracts can interact with them predictably, and for the token to be considered ERC–20 compliant. These functions, called methods in the ERC, describe what must be included in the smart–contract enabled token, while events describe an action to be executed.

Required Functions:

ERC–20 defines the following core functions (methods) that focus on tracking balances and permissions.

totalsupply = the total number of tokens in existence

balanceOf (address account) = returns the balance of a given address (the owner’s account)

transfer (address to, uint256 amount) = moves tokens directly from the caller to another address and must return a boolean indicating success, while also emitting a “transfer” event.

transferFrom (address from, address to, uint256 amount) = automatically executes the transfer of the specified number of tokens, from a specified address, from the approved spender, decrementing the token allowance

approve (address spender, uint256 amount) = lets a token owner authorize another address to spend up to a specific amount on their behalf

allowance (address owner, address spender) = how much is currently approved

Required Events:

“Transfer” and “Approval” events need to be emitted every time balances or allowances change, including mint and burn operations. Wallets and block explorers rely heavily on these events to reconstruct balance history and token activity, so skipping them, even in edge cases like zero–value transfers, breaks compatibility.

Transfer = an event triggered when a transfer is successful

Approval = a log of an approved event

Optional (but expected) Functions:

The following functions are optional and are not required, but they enhance the token’s usability, and virtually every consumer of ERC–20 tokens assume they exist. These functions should return token metadata, usually stored as contract state variables or exposed through simple getter functions.

name () = name of the token

symbol () = symbol representing the token

decimals () = eighteen decimals is the near–universal convention, matching Ether’s own denomination, though stablecoins sometimes use six.

Implementation Details

Ok, now we are moving into coding best practices and reliability techniques, which is way beyond the scope of the article. That said, we can illuminate that many ERC–20 implementations leverage existing software libraries that provide pre–configured core ERC–20 behaviors and the interfaces needed for a compliant fungible token, including things like ready–made extensions and presets for minting, pausing, burning, and access control patterns.

This accelerates implementation by reducing mistakes by providing audited building blocks rather than making developers build a custom token contract from scratch. These libraries provide a secure, well–tested implementation library for ERC–20 tokens, so developers usually inherit from its contracts instead of writing the standard token logic from scratch.

Popular supporting libraries/frameworks for ERC–20 work include OpenZeppelin Contracts, TokenLib and the ERC–20 Token Standard Library in the Meta Contract framework.

• OpenZeppelin Contracts: the dominant choice for building secure standard ERC–20 implementations and extensions like pausing, minting, and access control.

• TokenLib: a reference implementation library for Ethereum tokens, aimed at reusable token logic.

• ERC–20 Token Standard Library: a reusable ERC–20 library built for the Meta Contract framework, with core implementation and storage helpers.

• Utility libraries: rather than specific token frameworks, these libraries offer things like helper contracts for safe transfers, and access controls.

In practice, OpenZeppelin remains the most popular and is a default choice when people want the safest and most widely adopted path for helping to streamline building their ERC–20 implementations.

Gotchas

The biggest issue with ERC–20 creation is mismatching the expected return values or failing to emit events exactly when the standard expects them. Another common problem is handling allowances incorrectly, especially around changing an existing approval, which can create integration problems for external apps. Also note, that ERC–20 does not require a receive–hook for token recipients, so contracts that cannot handle incoming tokens may not be compatible at the standard level.

ERC–20 Benefits

ERC–20 remains popular because it lowers integration friction for developers and users. A token that follows the standard can plug into wallets, explorers, exchanges, and smart contracts with minimal custom work, instead of requiring one-off support everywhere. ERC-20’s main promise is that one token implementation can “plug into” a large existing Ethereum ecosystem. That standardization is the core reason it became so widely used.

• Interoperability: ERC-20 tokens are designed to work consistently across wallets, exchanges, and decentralized applications.

• Composability: ERC-20 tokens all expose the same predictable methods, so developers can build applications that plug into any compliant token automatically, which is often pointed to as a key enabler of decentralized finance and for decentralized applications (dApps) that require interaction with different tokens.

• Simplicity: Creating tokens takes minimal development time. For instance, startups launch their tokens to attract investments through ICOs or for loyalty programs. Developers creating a new token don’t need to design a transfer and accounting system from scratch.

• Integration: Because ERC-20 defines a common interface, any wallet, exchange, or smart contract that understands the standard can interact with any ERC-20 token without needing custom integration. ERC–20 tokens are used in DeFi protocols for lending and trading, as well as in the NFT ecosystem for auxiliary operations.

• Liquidity: DeFi, crypto exchanges and liquidity pools are built to recognize standard interfaces, so ERC-20 tokens can generally be listed and traded more easily than nonstandard tokens, giving new projects faster access to markets and users.

• Transparency: Blockchain allows tracking token transactions, ensuring trust for users and investors. Liquidity and market access are frequently mentioned as well.

• Security: ERC-20 has existed for years and been implemented, tested, and audited many times over, making it safer and more secure to use than inventing a new token mechanism, and because the ERC-20 common failure modes are well understood, they have well understood mitigation mechanisms.

• Network Effects: ERC-20 has clearly fostered network effects in the Ethereum ecosystem. Any new token or application that adopts it inherits a mature surrounding ecosystem rather than having to bootstrap its own tooling and supplemental support resources.

ERC–20 Pain Points and Risks

ERC–20 is not perfect and naturally inherits common pain points from the Ethereum block chain operations and infrastructure.

High Fees. Ethereum uses a “gas fee” model that depends on the current load on the network. During periods of high activity, fees for simple ERC–20 transactions can become expensive, especially for small transfers. Costs can spike unpredictably, which is a major issue for apps that need stable pricing.

Vulnerabilities. ERC–20 smart contracts require thorough testing. As with all crypto code, errors can render tokens inaccessible or vulnerable to hackers. Projects that do not undergo security and performance audits often become targets of attacks, leading to loss of funds. See the vulnerabilities section below for more details.

Scalability. The Ethereum network has a limited transaction processing speed (~30 TPS). With a high number of users, this can lead to delays and increased fees. Popular applications can crowd out other apps/users, causing delays and unpredictable confirmation times. Deploying onto Layer 2 chains is a common method for addressing these issues.

ERC–20 Implementation Vulnerabilities

As with any software, implementing secure ERC–20 contracts requires following hardened coding patterns, careful access control, upgradeability by design, exhaustive testing, executing security audits, and implementing operational safeguards (pausing, monitoring, key management). Below we outline a few areas of vulnerability specific to ERC–20. This is by no means a comprehensive list.

• Reentrancy Attacks: Attackers can repeatedly call the transfer function before the initial transaction is completed, potentially draining the contract’s funds. Implement ReentrancyGuard or similar and avoid external calls before state changes.

• Gas Limit Issues: If a transfer transaction runs out of gas, it may fail, causing users to lose gas fees without transferring tokens.

• Unlimited Allowance Exploit: Setting an unlimited allowance allows the spender to repeatedly drain tokens from the owner’s account without further consent.

• Incorrect Approval(s): Users might mistakenly approve a malicious contract, leading to unauthorized transfers when transferFrom is called. Minimize privileged roles and use role–based frameworks to separate powers for mint, burn, pause, upgrade into distinct roles and require multi–step access/governance for high–risk operations.

• Race Conditions: If the owner changes the allowance while the spender uses transferFrom, this could lead to unexpected or duplicate transactions. To counter this, implement a pausability/emergency–stop to freeze token behavior during incidents, with strict governance over who can pause/unpause.

• Incorrect Assumptions: Users might wrongly assume that allowance values update after a transferFrom operation, potentially leading to logic errors.

• Double Spending (Approval Race Condition): If the spender calls transferFrom during an allowance change, they could spend more tokens than intended. It’s safer to reset the allowance to zero before setting a new value.

ERC–20 Token Reception Issue

This issue has resulted in a large $ quantity of tokens being lost in the past. The ERC–20 documentation specifically calls out this issue. ERC–20 implementations are prone to this issue unless restrictions are implemented.

When ERC–20 tokens are sent to a smart contract that is not designed to handle ERC–20 tokens, those tokens can be permanently lost. This happens because the receiving contract does not have the functionality to recognize or respond to the incoming tokens, and there’s no mechanism in the ERC–20 standard to notify the receiving contract about the incoming tokens.

There are several coding tactics that can be used to help prevent the loss of tokens in this situation. One example: add explicit recipient checks in transfer flows, use “approve” + “transferFrom” for contract deposits. Other tactics (transfer rejection, extraction/rescue function, token–transfer wrappers, receiver notifications) can be implemented depending on the nature of the token use case.

Primary Use Cases

ERC–20 vaulted in popularity because it lowered integration friction for developers and enabled consistent, seamless, access for users. A token that follows the standard can plug into wallets, explorers, exchanges, and smart contracts with minimal custom work. The scope of use cases supported by ERC–20 token is already diverse and varied, and we can expect this eco–system to continue to expand.

• Stablecoins–are the highest volume category by far. The 3 largest US dollar pegged stablecoins – USDT, USDC, and DAI process hundreds of billions of dollars in settlement volume monthly across the Ethereum core network (called mainnet) and multiple Layer 2 networks. ERC–20 is the dominant format for dollar–pegged tokens because they need broad transferability and exchange support. Additional well known crypto tokens leveraging the ERC–20 standard include Binance USD (BUSD), Bitfinex LEO (LEO), BNB (BNB), HEX (HEX), Maker (MKR) and Shiba Inu (SHIB) to name just a few.

• Decentralized Finance (DeFi)–infrastructure is built almost entirely on ERC–20 tokens. Tokens in this space support asset swapping, lending, governance layer /voting, decentralized exchanges, and liquidity pooling.

  Example: Uniswap, is a decentralized exchange that lets users swap tokens directly onchain from their wallets without using a central intermediary. It has processed over $2T in cumulative volume, with its UNI token serving as the governance layer where holders of UNI can propose and vote on protocol changes, fee settings, treasury decisions, and other upgrades, which helps decentralize control of the platform.

• Payments and Transfers–Because ERC–20 tokens are standardized and widely supported; they are commonly used for simple value transfer and payment settlement within applications. Peer-to-Peer (P2P) payments in crypto are direct transfers of value between two parties that occur without a traditional financial intermediary; they can happen on-chain (direct wallet-to-wallet) or via P2P marketplaces.

  In P2P marketplaces/trading platforms: buyers and sellers connect on a platform that matches counterparties and typically holds the seller’s crypto in escrow until the buyer completes off-chain payment (bank transfer, payment app, cash), at which point the platform releases crypto to the buyer.

• Utility and Governance–is another major category. Utility tokens provide access to a specific product or service within a Distributed Application (d’App) ecosystem. d’Apps are composed of a network of components that are orchestrated to deliver services. Typical d’App components include a blockchain, smart contracts, wallets, oracles, users, and the surrounding protocols that connect them.

  Example: Chainlink’s LINK d’App is used to pay node operators for providing external data to smart contracts.

• Governance tokens grant holders’ voting rights in a project’s Decentralized Autonomous Organization (DAO), allowing them to influence the protocol’s future direction, fee structures, and upgrades. Projects like Uniswap (noted above) and MakerDAO use governance tokens to involve their communities in shaping the future of their platforms, fostering transparency and collaboration.

• Wrapped Tokens–allow assets from other blockchains to participate in Ethereum’s ecosystem. The most prominent example is Wrapped Bitcoin (WBTC), which allows Bitcoin liquidity to be utilized within Ethereum’s DeFi ecosystem.

• Real–World Asset (RWA) Tokenization–there is a massive amount of activity in this area of the crypto space. Ethereum hosts the majority of tokenized Real World Assets. RWA tokenization enables investors to acquire, hold, and sell tokens representing fractional ownership of real world assets such as government Treasuries, Bonds, Real Estate, and many other assets. The scope of physical assets being tokenized is exploding exponentially.

What is RWA Tokenization?

RWA tokenization is the process of representing ownership of a traditional financial or physical asset as a digital token on a blockchain. Rather than holding a paper title, a fund share, or a bond certificate, an investor holds a token that represents a fractional or full claim on the underlying asset, with the blockchain serving as the record of ownership

Total distributed asset value in RWA, reached approximately $34B as of spring 2026, up from approximately $5.4B at the start of 2025. Types of Physical/Real Assets being Tokenized:

• Treasuries and money market funds. This is the anchor category of the space with some of the largest and well know financial service providers offering tokens for Treasuries, ETFs, and other Money Market Funds.

• Private credit. This has become one of the largest categories by onchain value, covering direct lending and asset–backed credit structures and provides faster origination and settlement.

• Commodities. Primarily gold, and silver with the potential for other metals to come.

• Equities and ETFs. The tokenized equity market reached approximately $1B in early 2026. This category is gaining institutional legitimacy quickly: the SEC approved a NASDAQ rule change in March 2026 enabling tokenized Russell 1000 securities and major ETFs to trade on the exchange, with tokenized shares fully fungible with traditional shares on the same order book.

• Real Estate. A number of platforms focus on fractional property tokens, and some issuers distribute rental income to token holders in stablecoins.

• Bonds and Corporate/Sovereign Debt. Traditional fixed–income debt issuance is increasingly moving onchain, appealing to issuers partly on cost and 24x7x365 flexibility offered by the crypto/blockchain ecosystem

• Staking and Yield Farming. Users can lock their ERC–20 tokens in a protocol to earn rewards, this is the process known as staking. Yield farming involves earning additional tokens by providing liquidity to decentralized exchanges (DEXs).

Evolution

It goes without saying, the scope and range of standards for crypto tokens is evolving rapidly. ERC-20s well understood limitations is helping to foster the emergence of more specialized standards. A quick search showed me more that 20 new token standards are under active development such as ERC-1400, ERC-3643, ERC-1155, ERC-4337, to name just 4.

Another notable development is the emergence of pERC-20, a recently proposed Ethereum token standard, formally tracked as ERC-7605. pERC-20 will address the “lack of privacy” inherent in ERC-20, where today every wallet holding tokens has a publicly readable balance–anyone can query balanceOf on any address and see exactly how many tokens it holds, where they came from, and where they went.

pERC-20 proposes (it is not a live standard yet) making token transfers private by default, hiding balances, transaction amounts, and counterparties using zero-knowledge proofs baked directly into the token contract.

Wrapping up

New specialty token standards aside, ERC–20 remains the undisputed standard for fungible tokens on Ethereum and across all EVM–compatible chains. Its simplicity, widespread adoption, and the massive network effect of tools, wallets, and DeFi protocols built around it are making it a core foundation in the infrastructure to support today’s decentralized crypto world.

This is an example of how a well thought out standard can develop staying power based on network effects: every new wallet, exchange, and protocol built to support ERC–20 makes every existing ERC–20 token more valuable and accessible, a self–reinforcing dynamic that newer competing standards have found extremely difficult to disrupt.

Please, share your thoughts via the Comments section for this post or open a new SubStack chat thread … and please forward this post to your friends and colleagues. Until next time!

Let’s get started!

QKD is built on the foundations of quantum mechanics; notably quantum superposition, entanglement and the no–cloning theorem. This means, when QKD is correctly implemented, one would need to violate fundamental physical principles of quantum mechanics (kinda hard to do) to break a quantum key. For details on quantum superposition, entanglement and related topics, visit this Techaptitude post.

Traditional public key cryptography relies on the computational difficulty of certain mathematical functions, which are strong within the context of traditional computing technologies, but are highly suppliable to being cracked by quantum systems in the not too distant future.

Image Credit: In a High-tech Command Center, Professionals Analyze Complex Data on Multiple Screens by Aleksei Gorodenkov from Noun Project (CC BY-NC-ND 2.0)

A key defining property of QKD is the ability of the two communicating parties to detect the presence of any third party trying to gain knowledge of the key. This occurs due to the foundational characteristic of quantum mechanics where any attempt to measure a quantum system disturbs the system introducing detectable anomalies, thereby revealing the presence of the eavesdropper. This unique property ensures that the distributed keys remain secure, as any attempt at interception will be immediately apparent and will invalidate the exchanged key.

This domain is seeing a lot of research and commercialization work. The development of QKD systems has reached a high level of technical maturity, with multiple commercial vendors producing products tailored for various applications. New protocols, integration with Post Quantum Cryptography (PQC), and practical deployment over fiber and satellite networks characterize recent progress. Further, researchers are working on real–world implementations, multi–user systems, and addressing scalability, noise resistance, and security vulnerabilities.

Why We Need it

Today’s communications networks are vulnerable and subject to hacking and intrusion, and these risks are expected to grow exponentially from advances in quantum computing which are predicted to break all existing cryptographic systems in the 2030–2035 time frame; the so called “Q-Day”. See the discussion about Q-Day in this post on Post Quantum Standards.

Quantum computers will potentially render conventional public–key cryptographic scheme in use today like RSA (used for digital signatures, securing the Internet, public–private key exchange) and elliptic curve cryptography (used to secure crypto currencies, stable coins etc.) effectively obsolete. For more details on ECDSC’s role in Cryptocurrencies–see this article.

This threat is not purely hypothetical: it is strongly suspected that bad actors are already harvesting encrypted data today with the intention of decrypting it once sufficiently powerful quantum hardware arrives–a strategy known as “harvest now, decrypt later.”

Quantum Key Distribution addresses this vulnerability at a fundamental level by leveraging the laws of quantum mechanics to securely exchange cryptographic keys between two parties. QKD holds out the promises of:

• offers unbreakable security–including against future quantum computers, it holds even against an adversary with unlimited resources

• provides forward secrecy–keys generated via QKD cannot be retroactively broken regardless of how technology advances

• the inherent eavesdropping detection mechanisms enables real–time awareness of interception attempts, giving network operators an active rather than passive security posture

• QKD can provide protection against harvest now, decrypt later attacks

• offers long–term secrecy for sensitive communications–governments, defense, finance, telecom, and critical infrastructure operators are especially interested because they handle long lived and/or highly sensitive data

• has the potential to work over existing fiber optic links/networks

QKD vs. PQC

The QKD process should not be confused with Post Quantum Cryptography (PQC). QKD and PQC both aim to protect data from quantum era threats, but they do so in very different ways and we can expect both to become foundations for next generation security systems.

QKD uses quantum physics to distribute keys, while PQC uses new cryptographic algorithms designed to resist quantum attacks. In practice, PQC tends to be easier to deploy at scale, while QKD is more specialized and hardware dependent. QKD provides key exchange over a physical channel and can detect eavesdropping, but it needs dedicated networking hardware and is limited by distance and infrastructure constraints. PQC is software based, so it can usually be adopted within existing network stacks with far less hardware change. QKD is most attractive for fixed, high–security links such as government, military, or critical backbone networks. PQC is better suited to broad deployment across consumer, enterprise, and internet–scale systems.

Core QKD Protocols

There are several different methodologies for quantum key distribution. The BB84 protocol is a seminal QKD method first introduced in 1984 and is named for its inventors Charles Bennett and Gilles Brassard. BB84 is an example of a “prepare–and–measure” QKD protocol, where Party A performs the encoding by preparing the quantum states, Party A then forwards the quantum states to Party B, Party B receives and measures the states.

BB84 is the most established and is a simple QKD protocol, relying on single photons transmitted through a fiber optic cable, with polarization encoding, and each photon representing a bit of data (zero or one). Polarizing filters on the sender’s side set each photon’s orientation, while the receiver uses beam splitters to read the orientation. The sender and receiver then compare their photon orientations, with the matching set becoming the cryptographic key.

The parties exploit two communication channels — an insecure quantum channel and an authenticated classical channel. The quantum channel transmits signal states that an eavesdropper has complete access to.

In BB84 the measurement orientation is called “basis” and is combined with a bit value (state) into a qubit. The sender, Party A, prepares quantum states by randomly choosing a basis (z or x) and a state (either 0 or 1) then sends the qubit to the receiver, Party B, through the quantum channel–this is a completely untrusted channel.

Upon receiving the qubit, Party B also independently chooses a basis for measurement. If A and B both select the same basis, and there’s no interference or loss in the channel, they will obtain the same measurement result. However, if their chosen bases do not align, quantum mechanics dictates that the measurement outcomes are completely uncorrelated–so those events are simply discarded.

After transmission, to check for eavesdropping, Parties A and B publicly compare a sample of bits from that key to assess which basis was used for each photon (not the bit values) and keep only the results where their bases happened to match–discarding about half the data where mismatches have occurred. A significant mismatch indicates interference and the key is aborted. The shared measurement that matches becomes their shared secret key.

A bad actor or other 3^rd party can intercept the qubit and try to measure it, but must guess a basis, and if the guess is wrong, the measurement disturbs the qubit, and Party B will receive a changed version. Even simple attacks where qubits are intercepted and resent leave a distinct mark on the transmission which is easily detected.

What about just copying the qubits and waiting? This also interferes with the transmission and because of the no–cloning theorem, it is impossible to create an exact copy of an arbitrary unknown quantum state from a non–orthogonal set of states.

Non–Orthogonal Quantum States

The security of BB84 is based on the principle that two non–orthogonal quantum states cannot be perfectly distinguished. Stated another way, a non–orthogonal quantum state is a quantum state where it is difficult to distinguish one state from another. In practice, this means no measurement can tell them apart with 100% certainty in a single shot. This uncertainly means that the states cannot be reliably copied without disturbing the signal.

Additional QKD Protocols (this is only a partial list)

Prepare and Measure Protocols: (P&M)

This is largest classification of protocols, and as noted above, BB84 is the founding protocol of the field. Additional P&M protocols include the following:

• SARG04 – Researchers built SARG04 when they noticed that by using the four states of BB84, instead of 2, with different information encoding they could develop a new protocol which would be more robust, essentially a hardened version of BB84 with improved resistance to Photon–Number Splitting (PNS) attacks where an eavesdropper can take one photon from a multi–photon pulse and measure it without disturbing the communication.

• B92–is a simplified two–state prepare and measure protocol that transmits two non–orthogonal quantum states, the receiver randomly measures each incoming photon using one of two complementary bases, and when the measurement basis is “compatible” with the sender’s basis the result is unambiguous.

• MSZ96–uses four non–orthogonal quantum states of a weak optical field to encode key bits, explicitly avoiding photon polarization or entangled photons.

• The Six–State protocol–extends BB84 by using a six–state polarization scheme across three orthogonal bases, improving noise tolerance and error detection.

• The DPS protocol–is known for its simplicity and efficiency that removes the need for basis selection, uses sequential coherent pulses and has a simpler receiver configuration with fewer detectors, and is robust against photon–number splitting attacks.

• The COW protocol–transmits key information using weak coherent pulses, requires only a random number generator on the sender’s side, and achieves a high key transmission rate.

• The KMB09 protocol–increases transmission distances by using two mutually unbiased bases, making it particularly effective for higher–dimensional photon states.

Entanglement Based Protocols:

These methods distribute entangled photon pairs and leverage the properties of quantum entangled states to securely share cryptographic keys between a sender and a receiving party. A source produces entangled pairs (usually photons in Bell states) and sends one particle to Party A and one to Party B. When the parties choose compatible basis’s their outcomes are strongly correlated (or anti–correlated) and can be converted into raw key bits.

Entanglement protocols provide intrinsic eavesdropper detection through correlation/Bell tests, useful for device–independent or semi device–independent security proofs. Challenges include the need to generate high rate, high fidelity entangled pairs; transmission loss and decoherence can limit practical distance (entanglement degrades over fiber/atmosphere) requiring either trusted sources, quantum repeaters, or entanglement swapping to support long distances.

• E91–uniquely leverages quantum entanglement and Bell inequality violations to detect eavesdropping, giving it device independent security not reliant on trust in devices, but it has more implementation complexity and lower key rates.

• BBM92–is an entanglement based variant of BB84 using polarized entangled photon pairs together with decoy states to securely transmit non–orthogonal quantum signals. It provides equivalent key rates but simplifies some security proofs relative to BB84.

Device–Independent Quantum Key Distribution (DIQKD)

In ordinary QKD, there is an assumption that the hardware devices operate as specified. In reality, QKD systems are vulnerable to serious attacks, including side–channel attacks and detector loopholes. Stated another way, if the physical equipment in QKD is buggy, hacked, or even just slightly mis –calibrated, the security guarantee can break down.

In DIQKD, the physical devices are treated as potentially faulty or even malicious. In DIQKD we only trust the statistics the devices produce. The core idea is to verify security from observed measurement results, especially violations of a Bell inequality, rather than from a detailed model of the devices.

DIQKD is “trust the physics, not the box”. If the measurement correlations are strong enough to violate a Bell inequality, that is evidence the devices are behaving in a genuinely quantum way that can support secure key generation. This makes DIQKD a strong answer to side–channel and detector loophole problems that affect real systems.

Bell Inequality

As noted above, the key statistical test used in DIQKD is called a Bell inequality violation. When physicists measure the polarization of entangled photon pairs along various angles and compare the correlations, quantum mechanics predicts correlations stronger than any correlation a hidden–variable theory could produce.

This is the Bell violation–the correlation numbers in quantum mechanics come out higher than correlations based on hidden variables attached to the particles. This violation proves that the measurement outcomes were not predetermined by any hidden variables and that the two particles share non–classical correlations, leading Parties A and B to be confident that a secret key can be extracted.

DIQKD’s main challenge is it harder to implement than standard QKD because it needs very high–quality entanglement, low noise, and strong detection efficiency to close loopholes well enough for a secure Bell test. As a result, DIQKD usually has lower rates and is more technically demanding than measurement device–independent QKD, which is a more practical middle ground for many deployments.

Twin–Field Quantum Key Distribution (TFQKD)

Conventional (P&M) QKD faces a fundamental rate–distance trade off where the exponential decay of key generation rates with transmission distance due to optical fiber losses creates a serious limitation on the transmission distances that can be achieved. In traditional QKD this decay has been addressed by inserting physically secured relay nodes–aka add more hardware.

Introduced in 2018, Twin–Field Quantum Key Distribution (TFQKD) is designed to improve security and overcome this fundamental rate–distance limit without requiring trusted repeaters. TFQKD exploits single photon interference to enable quantum secured communication over unprecedented distances. The TFQKD protocol effectively uses phase locked signals from distant quantum light sources.

TFQDK has progressed quickly from early proof–of–principle demonstrations in 2019 to a 1,002 km fiber record in 2023 and the first free space demonstration in 2025. There are a number of remaining challenges; satellite integration, scalable multi–user network architectures, photonic chip deployment, and full composable security under realistic conditions are areas of active research. That said, TFQKD is widely regarded as the leading candidate protocol for long haul backbone links in the future quantum internet.

Real World Implementations/Status

The P&M QKD paradigm is the dominant approach in deployed systems because it is operationally simpler than entanglement based schemes and avoids the overhead of generating and distributing entangled photon pairs.

P&M QKD has graduated well beyond laboratory experiments and proof of concepts to where there are commercial deployments in fiber networks in many countries. Commercially available systems are provided by a number of vendors including Toshiba, ID Quantique, QRate, and QuantumCTek. There are metro scale deployments in China, Japan, South Korea, Switzerland, and parts of Europe.

All that said, today’s deployments tend to be specialized links and pilot networks, often subsidized by government programs or serving niche high–security needs. Satellite and long reach demonstrations have implemented ground to space links and trusted node satellite relays, proving feasibility of key distribution under controlled conditions.

Device–Independent QKD and Twin‑Field QKD can best be characterized as experimental/research lab based as both are active research areas with laboratory and proof–of–concept demonstrations but are not yet widely deployed commercial systems. Some observers would characterize these systems as limited–capacity prototypes. Most observers suggest DIQKD is at a much earlier stage than TFQKD.

DIQKD relies on loophole free Bell violations and therefore requires very high quality entanglement, near ideal detection efficiency, and distance/loophole control. Only a few proof of concepts and short range experiments have demonstrated device–independent key distribution and experiments have surfaced limitations related to transmission distance, key generation rate, and real world practicality.

TFQKD was designed to achieve a key rate higher than the maximum allowed for ordinary direct transmission over a lossy channel without needing full quantum repeaters. TFQKD has seen numerous lab–based proof of concept and long–fiber experiments, including implementations that beat repeaterless bounds and demonstrations over hundreds to >1000 km of optical fiber. TFQKD remains in the research/field trial stage and has not achieved scalable commercial deployment.

Standards Development

On the standardization front, the International Organization for Standardization (ISO), and the International Electrotechnical Commission (IEC) standards bodies are actively progressing to define QKD system interfaces, testing, and conformance for P&M systems and key‑management integration. The European Telecommunications Standards Institute (ETSI) has developed common criteria protection profiles specifically for P&M QKD modules. Standards for certification of complete QKD ecosystems have not yet been established.

Interesting Milestones

• Key Generation Rate Milestones–P&M systems have seen dramatic improvements in key generation rates. Recent experiments with the BB84 protocol have concluded that the key generation rate reached 2.3 Mbps, a meaningful step up from earlier kilobit per second regimes. At shorter metropolitan distances, a 2023 experiment achieved a QKD bit rate record of 115.8 Mb/s at 10 km.

• Satellite Based QKD–China successfully launched the world’s first quantum satellite, Micius, and completed a 2,000 plus km fiber optic network. In March 2025, researchers reported the development of the world’s first quantum microsatellite, Jinan–1, and demonstrated real time satellite based QKD with multiple compact ground stations in China and South Africa, enabling encrypted communication between Beijing and Stellenbosch over a distance of 12,900 km.

• Fiber Distance Records–with TFQKD Chinese scientists exchanged keys across 1,002 km of optical fiber without any intermediate repeaters or trusted nodes.

• Hybrid QKD and PQC Integration–a notable commercial milestone arrived when Toshiba Europe launched a commercial QKD system that integrates Post Quantum Cryptography using NIST’s ML–KEM standard. Get more details on NIST’s quantum standards here. This hybrid approach is considered an important interim milestone during the transition period before full quantum repeater infrastructure matures.

Wrapping UP

P&M QKD is simultaneously the most mature and the most constrained branch of QKD. It has achieved genuine operational deployments, dramatic distance records via TFQKD variants, and real time intercontinental satellite QKD, and is now integrating with Post Quantum Cryptography for defense in depth.

Standardization and certification gaps; while standards work is underway, fully mature conformance and certification regimes for operational security assurance are still developing and this can result in slowing multi‑vendor integration and commercialization.

While the QKD field has outstanding challenges–notably trusted node vulnerabilities, hardware side channels, stable/secure quantum repeaters, incomplete security proofs, and high costs. These are active research fronts rather than intractable barriers and good progress is being made.

The next decade will likely see P&M protocols coexist with hybrid PQC approaches and gradually cede long haul roles to entanglement–based or device–independent designs, while remaining the workhorse of metropolitan and satellite deployments for the foreseeable future.

Let us know what you think. Please share your thoughts via the Comments section for this post or open a new SubStack chat thread … and please forward this post to your friends and colleagues. See you next time!

Let’s Get Started!

Automated machine-to-machine payments have been conceptualized for years. AI Agents can do many things autonomously; analyzing market data and assessing credit risk are examples – executing direct financial transactions between digital services typically required human authorization, and until recently, have been beyond AI Agent capabilities. The x402 Protocol is being developed to extend AI Agent capabilities into the automated real-time payments space.

x402 is an open standard for internet based payments. Developed by Coinbase, it enables AI Agents to execute payment transactions autonomously. X402 is a native implementation of HTTP that leverages blockchain and digital currencies (such as the USDT stablecoin) to deliver secure, lightweight, and instantaneous payments. It aims to support both crypto and fiat networks using a variety of types of value - stablecoins, blockchain tokens, fiat currencies, etc. x402 is compatible with any EVM –compatible blockchain with current documented implementations on the Base, Polygon, and Solana blockchains.

An ecosystem of vendors is evolving to support x402, and it currently includes American Express, AWS, Circle, Cloudflare, Coinbase, Google, Mastercard, Stripe, Shopify, and Visa. Implementations are currently executing stress and security testing with an active and growing open-source community.

New On-Demand Payment Mechanisms

The underlying rationale for why x402 is needed is simple. AI agents operating across commerce, financial services, data analytics, business and licensing relationships and will need dedicated payment mechanisms, rules, and protocols that remove the need to have humans authorize every transaction. Here are a few of the ways x402 can be put to use:

• Agentic (automated) Payments – enables AI agents to pay on-demand in real time and only pay for relevant content/assets aligned to the user’s needs/wants without human intervention. Buyers can execute pay-per-news article, pay-per-second viewing of video, pay-per-data look up, while eliminating the need for paywalls, account and subscription set up before they can purchase assets.

• Micropayments and Seamless Pay-per-Use – makes micropayments economically viable for merchants. A research journal enables users to pay-per white paper vs. requiring an annual subscription, a gaming platform charges per minute of play vs. requiring the user to pre-purchase large blocks of time, a pod cast charges per episode vs. requiring an annual subscription.

• Machine to machine (M2M) transactions – execute high velocity financial exchanges between automated systems, such as Internet of Things (IoT) devices, and purchasing cloud infrastructure on-demand.

• Simplicity – merchants don’t need to build or maintain blockchain infrastructure. x402 handles payment verification and settlement so that sellers don’t need to maintain their own infrastructure.

• Flexible billing schemes – supports multiple payment “schemes” > 402 currently supports “exact” (transfers a specific $ amount), “upto” (caps payment to maximum amount), and “batch-settlement” (roll multiple small charges into one transaction). Additional payment schemes will likely be added in the future.

• Streamlining Payment Operations –faster access, and the elimination of friction for buyers and faster settlement and cash flow for sellers, with less operational overhead on both sides.

How x402 works …

A simplified x402 transaction flow is outlined below. When contacted by a client, a web based service returns a 402 response with payment instructions, and clients then pay programmatically without accounts, sessions, or API keys or human intervention.

Simplified Transaction Flow Step by Step:

• A client attempts to access a gated web based resource (payment required) by x402.

• The server responds with a 402 PAYMENT-REQUIRED header containing Base64-encoded payment details: the price, accepted token, network, and merchant address.

• The client constructs a signed payment payload and retries the request with a PAYMENT-SIGNATURE header – aka the client re-requests the gated resource, this time with the payment authorization header.

• The server verifies the payment payload – directly or by calling a facilitator – and if validated, settles the transaction on-chain. The payment facilitator verifies the client’s payment payload and settles the transaction.

• X402 currently supports the USDC and EURC stablecoins for payments. X402 supports any ERC-20 token on EVM-compatible networks as well as SPL tokens on the Solana blockchain.

• The server returns the resource with a PAYMENT-RESPONSE header containing settlement confirmation, and with the requested resource in the response.

This flow creates programmatic pay-per-access or pay-per-use models for resources across the Internet. Clients and servers capable of interpreting the x402 protocol can transact without the need for accounts, subscriptions, or API keys, AND most importantly, without needing human approval/authorization.

What is/how is a signed payment payload created?

Cryptographic signing, think of this as a digital signature, adds a super-secure electronic stamp to a message. Signing is unique to each message, and it can’t be copied or faked. This digital signature proves (authenticates) that the message is truly from you, and that the message has not been changed, altered, or tampered with (integrity).

In x402, a client uses its crypto wallet to cryptographically sign the payment message payload following the EIP-712 standard to create the cryptographic authorization for the specified stablecoin transfer.

This signing process is a classic implementation of public-private key pair cryptography – a process that is commonly used in digital security. Net-net, this signed message cryptographically proves the message from the client is legit and consents to a specific transfer, to a specific merchant, of a specific dollar amount. The Facilitator then executes the payment to an actual blockchain.

The Key Players

Client – is any entity that requests a paid resource: a human-operated app, an AI agent, or a programmatic service. Clients only need a crypto wallet – no accounts, credentials, or session tokens to manage.

Server – defines payment requirements in the x402 response, verifies incoming payment payloads, settles the transaction, and serves up the resource. The x402 SDKs and a Facilitator handle most of this automatically.

Facilitator – the facilitator is an optional but recommended third-party service that abstracts blockchain interactions. Rather than connecting to a node directly, the server delegates two operations:

• POST /verify – Confirms the client’s payment payload is valid before the server fulfills the request.

• POST /settle – Submits the verified payment transaction to the blockchain.

The Facilitator does not hold funds. It verifies and broadcasts the Client’s pre-signed transaction on behalf of the Server. Multiple Facilitators are available https://docs.x402.org/dev-tools/facilitators across different networks.

Payment Schemes and Networks

Schemes are a logical way of moving money on blockchains. To help facilitate an expanding number of payment use cases/types, the x402 protocol is extensible to different ways of settling payments via its “scheme field”. Each payment Scheme may have different operational functionality depending on what actions are necessary to fulfill the payment.

Currently supported payment Schemes include:

• exact: transfers a specific $ amount for a single request or resource, such as paying a fixed price to read an article.

• upto: authorizes a maximum amount per request, aka a price cap, while the seller settles the actual usage up to that cap.

• batch-settlement: designed to reduce blockchain overhead by letting sellers collect many small charges in batches instead of settling every 402 payment request separately.

The batch-settlement Scheme is important to support merchants for handling automated and API payments that may be frequent and low value. Batching lets a merchant collect many micro-charges and settle them together, which can make the payment flow economically viable at scale.

New Scheme currently under development: deferred: this Scheme has been proposed for use cases where immediate settlement is not ideal, such as agents and crawlers that need delayed settlement, aggregated accounting, subscriptions, or pre-negotiated licensing.

The supported number of Schemes in x402 has been intentionally kept small at the outset to help contain complexity, and note, the protocol is designed to accept new Schemes through its extensible architecture. The x402 payment scheme specifications are documented in the project’s open–source GitHub repository at https://docs.x402.org/schemes/overview

Networks

In the x402 protocol, a Network refers to the Blockchain (Base, Solana etc.) used for Scheme’s transactions and settlement. The relationship is that x402 pairs Schemes and Networks as (scheme,network) (scheme,network), because the same Scheme can be implemented differently on different chains. Clients and Facilitators must explicitly support different (scheme, network) pairs to be able to create proper payloads and verify/settle payments.

The same Scheme requires different implementation details on the Ethereum versus the Solana blockchain. In the real world, that means “exact on Ethereum” and “exact on Solana” are the same in the functionality they execute and the result they create, but they require different technical implementations.

Security

The nitty gritty details of security in x402 are beyond the scope of this article. That said, here is surface level overview of the core aspects of the layered approach to security in x402. Many of the underlying security techniques and technologies used in x402 are well known and currently exist in HTTP, blockchain, and API systems. x402 implements them in specific ways to secure 402 based transactions.

• Signed buyer authorization – the client returns a signed payment payload in headers such as PAYMENT-SIGNATURE or X-Payment-style fields depending on the implementation. The buyer signs locally in its own runtime, and that signature is the key security primitive, so if anything is altered in transit, verification fails.

• Facilitator-based verification and settlement – x402 implements a facilitator to verify the payment payload and submits settlement on behalf of the seller – aka it only checks signed payloads and executes settlement logic.

• On-chain settlement – settlement happens on-chain, so the final payment record is cryptographically anchored in the chain rather than depending on an off-chain ledger controlled by the merchant.

• Replay and duplication defenses – payment proofs are intended for a specific request and should only be once, and use of a short-lived SettlementCache in some implementations can reject repeated settlement attempts for the same transaction payload.

• Authorization layer – it is recommended to implement policy-based access controls so that only approved roles, attributes, or relationships can access paid data after verification.

• Policy/risk controls for AI agents (optional)– ecosystem tooling can add pre–payment risk checks, traces, and fraud/dispute evidence to the standard flow, aiming to catch prompt injection, subscription traps, and similar agent–specific threats before settlement.

The x402 protocol is designed so the seller can verify payment without ever holding the buyer’s private key, which reduces custody risk and makes the flow stateless.

Test Implementations

Testing activities, both for security and performance/stress have ramped up quickly and are being driven by a combination of academic researchers, security firms, and x402 open-source community participants. Activity went from essentially nothing in mid-2025 to well over 100 million cumulative transactions through Q1 2026, with late 2025 seeing a surge of over 10,000% in transactions.

The protocol has attracted serious formal security scrutiny. A May 2026 paper from researchers at Ohio State, CSIRO, and the University of Manchester formally analyzed x402 and demonstrated that it is vulnerable in both design and implementation. The paper detailed five concrete attacks revealing weaknesses in authorization, binding, replay protection, and web–layer handling. The researchers validated these attacks through a reproducible testbed on local chains, Base Sepolia, and live endpoints, and audited three open-source SDKs, finding all five attacks were practical.

Formal stress testing under controlled conditions remains an area where publicly available documentation is sparse – the protocol architecture has not yet been subjected to a comprehensive and rigorous load-testing regime.

A couple of real-world security incidents have added urgency to the need for robust testing.

• 402Bridge hack, where 200 users’ USDC were stolen (with a total value of ~ $18K USD) due to permission mismanagement and poor private key custody: the contract lacked multi-signature mechanisms, all critical operations were controlled by a single admin key, this adini key was leaked, enabling the attacker to make off with the USDC stablecoins, and the project lacked regular security audits.

• March 2026, a signature verification bypass vulnerability was disclosed in the x402 SDK, a reminder that even a sound protocol design can carry implementation level flaws.

Wrapping Up

x402 introduces minimal extensions to well established and understood standard HTTP web flows and by maintaining full HTTP compatibility while weaving payment negotiation directly into the familiar request–response cycle, it doesn’t require a new protocol, new infrastructure, or accounts – it just adds a payment handshaking to the web’s existing plumbing that can be leveraged by AI Agents. All of this should help accelerate adoption and the growing list of vendors supporting the x402 open-source Foundation is another positive indicator for 402 adoption.

That said, nothing is a given in the currently turbulent land of AI, blockchain and crypto! Buyers/clients need to become comfortable with trusting their hard earned cash to AI agents, and blockchains and they must be willing to adopt crypto wallets. x402 payment systems rely on blockchain network availability, speed, and security. Limited awareness, aka fear of the unknown, compared to traditional payment methods, on both on the merchant and client/buyer sides could slow adoption big time.

Lastly and not surprisingly, regulatory uncertainty surrounding cryptocurrency based payments (and AI in general) can foster a lack of transparency will likely be a key inhibitor to adoption over the short/medium term.

The x402 open-source community has published a full reference implementation with core protocol libraries, client libraries for browser and Node.js environments and cryptographic utilities for signing and verifying payments.

Please, share your thoughts via the Comments section for this post or open a new SubStack chat thread … and please forward this post to your friends and colleagues. Until next time!

Let’s Get Started! **

Unlike most cryptocurrencies, which can often be subject to dramatic price swings (aka Bitcoin), Stablecoins are pegged 1:1 to less volatile assets such as fiat currencies (US Dollar, Euro etc.), crypto assets, or commodities (Gold). Stablecoins, which first appeared in 2014 attempt to combine the technological benefits of blockchain — transparency, efficiency, and programmability — with the financial stability needed to establish trust and to foster widespread adoption.

A primary goal of Stablecoins is to address the issue of crypto price volatility, and in doing so Stablecoins have unlocked new use cases beyond trading and speculation, appealing to a broad range of crypto users, both retail and institutional.

BitUSD, was the world’s first Stablecoin, released in July 2014, and was operated on the Bitshares blockchain. BitUSD was crypto-backed and not backed by the US Dollar.

In Sept of the same year, the NuBits Stablecoin was launched and was also pegged to the dollar, using the BitShaes cryptocurrency for collateral. NuBits relied on a flawed algorithm that couldn’t handle the extreme market volatility of the cryptocurrency market and the stablecoin lost its $1 peg in 2018 and never recovered.

In Nov 2014, Tether USDT was released and it has progressed to become the world’s most valuable and most traded Stablecoin with ~ $190B in Reserve Asset Base. Because fiat backed Stablecoins are pegged 1:1 to the US Dollar (USD), Reserve Asset Base (the total reserves backing the Stablecoin) is the best way to characterize a Stablecoin’s size and adoption.

Industry sources currently estimate there are ~ 380 Stablecoins in existence today – this includes dormant or low-volume coins. Conservative estimates have the number of active Stablecoins worldwide at ~180–200, with more than 99% of the Stablecoin supply being US Dollar denominated. The largest 2 Stablecoins (USDT + USDC) together represent ~86% of total market capitalization. Many algorithmic Stablecoins collapsed after TerraUSD failed in 2022, reducing the active count to ~ 20 worldwide. See below for more details on the TerraUSD collapse.

Stablecoin Designs – How They Work – Why that Matters!

Stablecoins are a special type of cryptocurrency designed to maintain a fixed value over time. Unlike volatile cryptocurrencies like Bitcoin and Ethereum, Stablecoins are “pegged” 1:1 to a traditional currency, most commonly the U.S. dollar, and they are backed “stable” asset(s) (Fiat Currency, Cryptocurrency, Gold), or a collection of stable assets.

The mechanism behind a Stablecoin type determines far more than how it holds its peg. It shapes how regulators classify it, what compliance obligations apply, where risk is concentrated and who bears responsibility when something goes wrong. The most common Stablecoin types are:

• Fiat currency backed: are backed by real-world fiat currency assets and currently dominate the market. Tether’s USDT and Circle’s USDC, the two largest Stablecoins, collectively account for more than $220 billion in market cap – fully 86% of today’s Stablecoin universe.

• Algorithmic: these coins rely on technical mechanisms to maintain the 1:1 peg. When the coin trades above its target, the protocol mints more supply; when it trades below mtarget, it reduces supply using redemption incentives to keep the price stable. These Stablecoins are not backed by collateral assets. This approach is riskier because it’s more susceptible to market fluctuations and technical failures.

• Crypto backed: Crypto-backed models not surprisingly are tied to cryptocurrencies, notably BitCoin and Ethereum which makes them more decentralized but also more complex and riskier. Smart contracts and overcollateralization are used to maintain their dollar peg.

• Commodity-backed: here the Stablecoin is pegged to tangible physical commodities such as precious Metals, Oil and Gas and/or Treasury Bills. The issuer typically holds reserves of the underlying physical commodity (like Gold). Commodity backed coins need to manage the variables of the market value of the underlying asset (aka the price of Gold fluctuates) and the physical infrastructure needed to store the assets.

Regulatory treatment follows these structural differences. Fiat-backed Stablecoins increasingly fall under dedicated payment stablecoin frameworks. Crypto-backed and algorithmic models are less clear-cut, with classification varying by jurisdiction and are still evolving. Commodity-backed tokens may face dual regulatory oversight.

Note on terminology used in this article. For simplicity, the words stablecoin, coin, and token are sometimes used interchangeably in this article. For true clarity, the three words describe a hierarchy in crypto assets: coin is the broadest term for a native cryptocurrency on a blockchain, token is a unit of value built on top of an existing blockchain, and stablecoin is a specific type of token.

Deep Dive into Stablecoin Types

Fiat Currency Backed Stablecoins

These are by far the most popular type of stablecoin — as noted above, they are tied 1:1 to the value of traditional currencies, with the USD and Euro being the most common benchmarks – surprise – surprise! These Stablecoins derive their stability from reserves held in the fiat currency or equivalent assets. Examples include the US Dollar pegged Tether USDT, Circle’s USDC, and the EUR–pegged Stasis Euro EURS.

Issuers of fiat-backed Stablecoins typically establish a reserve fund holding real-world assets, the volume of which mirror the total value of invested funds. So, for example, if a Stablecoin is backed by the U.S. dollar, the issuer will hold $100 million in US dollar assets to support 100 million Stablecoins issued to users. When a user wants to redeem their Stablecoin, the issuer can draw from this reserve to provide the equivalent amount of fiat currency to the user. Reversely, when a user invests in Stablecoin, the issuer acquires more of the collateral assets and deposits them into the reserve fund.

Example: Tether USDT

Tether launched its USDT Stablecoin in 2014, and in doing so essentially eliminated a lot of the hassle that comes with moving your money in and out of the crypto space, providing crypto investors with an exchange mechanism that is simpler to execute, and less volatile in price than buying crypto currencies directly. Prior to Tether, there really wasn’t any way for crypto traders to easily move their money in and out of the blockchain ecosystem without taking their fiat money out of the crypto space entirely.

This means if you wanted to sell your Bitcoin, for example, but didn’t want to cash out entirely and instead wanted to buy another cryptocurrency, you’d have to sell the crypto, take the gains in the form of actual cash, then re-transfer that cash into your crypto account, and purchase the new cryptocurrency. With Tether USDT, investors can sell their Bitcoin and then park their funds in a Stablecoin rather than cashing out of the blockchain.

Tether USDT coins are backed 1:1 by U.S. cash and other assets held in Tether’s reserves. Tether’s parent company, Tether Holdings Limited, manages the reserve supply. The company audits and reports on its reserves quarterly.

Tether says USDT is backed by reserves that include a large share of highly liquid assets, especially U.S. Treasury bills and other cash-like instruments. The collateral backing USDT is not just cash in a bank account; rather it is a mix of liquid assets and other holdings that together are intended to support the circulating supply.

Recent reporting cited by third parties suggests U.S. Treasuries and related short-duration instruments make up the bulk of USDT reserves, while more volatile assets, such as money market funds, secured loans, Bitcoin, precious metals, and corporate bonds may also be included in the reserves – al be it these are a small portion of the total reserves.

Here’s a closer look at how Tether stores its reserves in the order of ~ $160B in assets.

Tether’s liquid asset allocation (~ 84 percent) *

• U.S. Treasury bills: $94 billion

• Overnight USD reserve repurchase agreements: $17 billion

• Term reverse repurchase agreements: $8.5 billion

• Money market funds: $15 billion

• Cash and bank deposits: $400 million

• Non-U.S. Treasury bills: 100 million

Tether’s alternative asset allocation (~ 16 percent) *

• Secured loans: $8 billion

• Bitcoin: $7.5 billion

• Precious metals: $6 billion

• Other investments: $4 billion

• Corporate bonds: $10 million

* Tether’s asset mix changes dynamically both in terms of the types of assets and the relative investment quantities. The data presented above is just that, an example of the various collateral assets USDT leverages to collateralize the USDT Stablecoin at a given point in time. For more details visit this site: https://tether.to/en/transparency/?tab=usdt

Algorithmic/Synthetic Designs

Algorithmic Stablecoins are digital currencies that maintain their value through programmed mechanisms that adjust supply based on market demand, without relying on direct collateral > aka typically there are no hard assets (fiat currency or crypto) backing these types of Stablecoins.

These coins also target a $1 USD peg, and they deploy onchain rules and incentive systems to constantly adjust how many tokens exist. When the price drifts above a dollar, the system mints more tokens. When the price slips below a dollar, it removes tokens from circulation. These are self-correcting systems that keep the coin close to its peg value with adjustments happening automatically through smart contracts operating on blockchains. The typical adjustment mechanisms include:

• Rebasing systems – where an adjustment protocol changes the total number of tokens in circulation on a set schedule. If the price rises above $1, everyone’s balance increases; if it falls below $1, balances shrink. A user continues to hold the same percentage of the network, but the number of tokens they have shifts. This design makes each token more or less scarce until the market price returns to the target.

• Dual-token models – here the Stablecoin is paired with a second token that absorbs volatility. When the stablecoin is above $1, the protocol mints more of the secondary asset– sometimes distributing new tokens to holders of the secondary asset – to push the price down. When the price drifts below $1, users are incentivized to buy discounted stablecoins or swap the secondary token for stablecoins that get removed from circulation, tightening supply until the peg recovers.

• Hybrid or fractional models – these models blend collateral asset backing with algorithmic corrections. A portion of the Stablecoin’s value is backed by real assets, and the algorithm handles the rest. This mix can be more resilient because the collateral provides a stable buffer, and the algorithm keeps the coin responsive to demand. In theory, fractional models can maintain a tighter peg than fully unbacked systems, especially during volatile periods.

Algorithmic Stablecoins operationalize several tools and techniques including Smart Contracts which are automated software programs that execute the stablecoin’s monetary rules on the block chain(s) by creating, called minting, and removing, called burning, coins to manage supply changes, and coordinating swaps between paired assets.

Price Oracles are services that deliver real-time market prices to smart contracts on the blockchain, so the algorithm knows when the coin is trading above or below its target. There are number of price oracle types including Push-based price feeds that update a schedule or when price moves by a pre-set threshold, Pull-based price feeds that enable protocol to fetch high-frequency data when needed, Decentalized Exchange (DEX) based that derive prices from on-chain trading activity, often using time-weighted averages to reduce manipulation risk. Some Stablecoin designs use multi-source aggregated feeds that combine data from many exchanges or venues to avoid depending on a single market source.

Blockchain Networks are an obvious major critical component, as the chain’s speed, fees, and reliability shape how well the stabilization mechanism performs. Congestion or blockchain outages can delay transactions, limit arbitrage, and weaken the Stablecoin price by negatively impacting the peg.

Example: Ethena USDe

Ethena’s USDe, a synthetic US dollar pegged stablecoin, that uses crypto assets and automated hedging to maintain its dollar value, and it also operates the globally accessible dollar savings asset, sUSDe. USDe was founded by Guy Young in July of 2024 and released USDe in Feb of 2024.

It has a Reserve Asset Base of ~ $5.5B and operates on the Ethereum blockchain. The Stablecoin delta-hedges Bitcoin, Ethereum and other governance-approved spot crypto assets using perpetual and deliverable futures contracts, as well as holding liquid positions in the USDC (Circle) and USDT (Tether) Stablecoins.

USDe is decentralized and uses what’s called “cash-and-carry trade,” a kind of arbitrage that takes advantage of the price differences between long exposure and short derivatives exposure to Ethereum and some of its financial derivatives. USDe takes money client’s remit to buy USDe and uses that money to buy Ethereum (going long) while simultaneously betting against its price (going short) using financial derivatives.

This strategy attempts to neutralize the price risk of USDe, and it offers buyers a yield (aka interest) by staking the Ethereum and collecting payments from the short derivatives positions. In other words, USDe generates returns (yield) from its hedging activities and Ethereum staking activities and these returns are distrusted to the holders of USDe creating a passive income stream.

Governance and decision making for Ethena leverages the ENA governance token. Holders of the ENA Token can vote bi-annually to elect members to a Risk Committee, and in the future additional committees performing critical roles within the Ethena ecosystem.

TerraUSD Stablecoin Collapse

TerraUSD, once the third-largest stablecoin by reserve assets, is a stark example of the pitfalls of algorithmic Stablecoins. TerraUSD relied on a complex system of arbitrage and other cryptocurrencies to maintain its 1:1 peg to the U.S. dollar.

In 2022, a sudden crash eroded confidence in the system, leading to a catastrophic collapse with investors panicking and selling their TerraUSD tokens which broke the peg and it never recovered. The TerraUSD cost its investors $40 billion in losses and severely damaged the reputation of Algorithmic Stablecoins in the market.

Crypto Backed Stablecoins

As you might gather, crypto backed Stablecoins establish their peg by holding other cryptocurrencies as collateral rather than cash, money market funds, or government bonds. Given that the underlying collateral assets are volatile, these systems rely on a decentralized model with overcollateralization and smart contract rules to maintain their peg.

For example, if a Stablecoin has $10B of their Ethereum-backed stablecoin in circulation, they will hold more than $10B of Ethereum cryptocurrency in reserve to establish its overcollateralization which attempts to address the potential for volatility in the reserves.

Some of the common characteristics of these types of coins include:

• Overcollateralized vaults – users lock crypto into a smart contract and mint stablecoins against it, usually with collateral worth more than the coins issued.

• Decentralized governance – many crypto-backed systems are managed by protocols or Decentralized Autonomous Organization (DAOs) instead of a single company, with rules encoded in smart contracts and tokens used to manage membership in the DAO.

• Liquidation mechanisms – If the crypto collateral value falls too far, the system can automatically liquidate part of the position to protect the peg.

• Stability incentives – fees, interest-like charges, and arbitrage opportunities encourage users to keep the peg close to the target value.

• Redemption and mint/burn logic – tokens are created when collateral is deposited and removed from circulation when users repay or redeem them.

How Does a Decentralized Autonomous Organization (DAO) Operate? DAOs implement Stablecoin management via tokens and smart contracts to decide how Stablecoins are held, moved, invested, or used for payments. In practice, members vote on treasury policy, risk limits, yield strategies, and spending rules, instead of a single company or manager doing it alone

Example: DAI Stablecoin

DAI is a Stablecoin that uses cryptocurrency collateral locked in Maker Protocol smart contracts to maintain its 1:1 peg to the US dollar. DAI is a popular Stablecoin typically ranked in the top 5 globally based on its Reserve Asset Base of ~ $4.5-$5B.

Each DAI coin is backed by cryptocurrency and is overcollateralized by crypto deposits through Maker Protocol’s smart contracts, called Maker Vaults. This means that every coin requires a coin holder to lock cryptocurrencies worth more than the amount they mint in a Vault to ensure the stability of its $1 peg, even with fluctuations in the crypto market.

The Maker Protocol is open-source software built on Ethereum, and its Stablecoin can use a variety of ERC-20 tokens as collateral, including Ether (ETH), Basic Attention Token (BAT), wrapped Bitcoin (wBTC), Compound (COMP), and more.

The Maker Foundation, the company who coordinates the original development of the Maker Protocol, was co-founded in 2014 by Rune Christensen and Wouter Kampmann. Its first collateral token backed only by Ethereum was released in 2017, with multi-collateral functionality introduced in 2019. In July 2021, the Maker Foundation gave up control of the Maker software to MakerDAO, a Decentralized Autonomous Organization providing the protocol with a fully decentralized governance model.

How it works:

To create new coins, a user deposits their cryptocurrency (like Ethereum, ETH) into a Maker Vault. Their tokens are locked in the Vault’s smart contract as collateral against which they can withdraw, or more accurately, borrow DAI. The user’s deposits are referred to as a Collateralized Debt Position (CDP), and they must pay a small stability fee for using the service.

Maker uses pricing oracles to determine the dollar – based value of the deposited cryptocurrency, and users generate a specified amount of coins based on that value. This collateralization is never 1:1 because of the volatility of the cryptocurrency market. Therefore, Maker requires overcollateralization of each CDP in which users must deposit more value than they are taking out. If the dollar value of a user’s cryptocurrency decreases, then they must return the tokens they have borrowed, deposit more value into the Vault, or risk being liquidated.

Maker’s smart contracts employ several automated auctioning mechanisms to stabilize the cryptocurrency’s value around its peg to the US dollar. The auctions are implemented through smart contracts in the Maker Protocol that are executed autonomously without human intervention.

Maker also leverages something called Keepers, which are third-party participants (usually automated) that buy DAI when it is below its $1 target and sell it when it is above its $1 target. This process of arbitrage keeps a constant pressure on the price of the stablecoin to remain in a tight band around the Peg.

Commodity Backed Stablecoins

Commodity-backed Stablecoins are dominated by coins that are tied to the value of precious metals – Gold and Silver. The OIL1 Stablecoin was announced in January of 2026, with an expected release in the first half of the year. It will be collateralized by verified reserves of Gulf crude oil and pegged to both the U.S. dollar and the price of Gulf crude.

Commodity backed Stablecoins offer users the ability to gain exposure to commodities without directly owning them. Examples include:

• Tether Gold (XAUT) with ~$2.5–3.3B in gold reserves ~16.2 metric tons; 644 gold bars (7,667 kg) stored in Swiss bank vaults.

• PAX Gold (PAXG) with ~$1.6–2.3B in gold reserves, with each coin backed by one ounce of a 400-ounce gold bar stored in London based bank vaults.

• Kinesis Silver (KAG) – the largest silver backed coin, with ~$285–414M in Silver reserves, with 1 ounce of investment-grade silver bullion per coin stored in fully insured, audited vaults

Stablecoins - General Benefits

Stablecoins have become an essential tool in the cryptocurrency ecosystem, offering several key benefits.

• Reduced Transaction Fees: Many cryptocurrency exchanges skip the fees for users converting to or from Stablecoins. Instead of cashing out into U.S. dollars and racking up fees each time. International Transfers cost less than $1 vs. $10–$25 for traditional international wire transfers.

• Speed: Transactions settle almost instantly vs. days for traditional wire transfers and are available 24/7/365 (worldwide), aka transactions are unrestricted by traditional banking hours

• Hedging Against Volatility: By holding Stablecoins, traders can protect their investments from the price swings inherent in the crypto market and provide stability essential for corporate treasury and payments.

• Passive Income Opportunities: Some Stablecoins allow users to earn interest through staking or lending. For example, Coinbase offered a 4.1 percent reward to users who held USDC on the platform in July 2025.

• Programmability: Smart contracts and blockchain enable automated payments, yield lending, and collateral use.

General Risks

Anything in the crypto world has inherent risks.

• Depegging: If the value of the stablecoin deviates from underlying fiat currency (trades significantly below or above its $1 target), due to market conditions or liquidity issues, the issuer is forced to take corrective action quickly to restore confidence and the Peg. See pegging/depegging section below for additional details

• Reserve Risk: The Coin issuer may not hold the claimed collateral or reserves may be inaccessible. Recent regulation in the U.S. requiring audited reserves should help mitigate this risk.

• Counterparty Risk: Users need to trust that the Coin’s issuer is solvency and stable

• Regulatory Uncertainty: Accounting/treatment rules still evolving (FASB, IRS), and asset classification debates (SEC, CFTC, Treasury OCC)

• Technology Risk: Smart contract vulnerabilities, price oracle failures, network congestion can cause serious problems

• Liquidity Concentration: Most liquidity is concentrated in USD – backed stablecoins; other currencies are less supported

• Mass Redemption Risk: Unforeseen market forces may trigger large scale redemption requests which can overwhelm the issuer

• Transparency: Some Stablecoin issuers have been guilty of being less than transparent, given the ever shifting regulatory environment historically

Unique and Specific Risks:

Fiat-Backed Stablecoins: While generally recognized as the most risk-adverse type of Stablecoin, fiat-backed coins have risks related to centralization and single institution solvency, reserves may be frozen or inaccessible, and unclear/evolving regulatory overhang.

Algorithmic Stablecoins: These Stablecoins do not rely on backup collateral/assets, rather deploy automated supply adjustments via programmatic smart contracts to maintain the Peg. Algorithmic stablecoins depend on a coordinated set of technologies that let them run without a central operator. A loss of confidence can create a rapid downward spiral, and the lack of back up collateral means there is no hard asset floor – there is nothing to absorb losses if the peg breaks. Pure algorithmic models do not have a good historic track maintaining their peg.

Crypto-Backed Stablecoins: These Stablecoins are purposely over collateralized to help absorb volatility, yet, cryptocurrencies are notoriously volatile, so when used as backing collateral for Stablecoins, a crypto crash can trigger mass redemption events for the coin that can contribute to the possibility of a depegging situation.

Commodity-Backed Stablecoins: Typically backed by physical commodities – precious metals (Gold) and/or government issued Treasury Notes, these Stablecoins also have unique risks. Commodity prices do fluctuate which necessitate buying/selling assets by the issuer, and physical assets require securing storage and custody and therefore have lower liquidity – it takes time to buy/sell these assets.

Pegging and … Oh No … Depegging!

In a generic sense, a “peg” is a specified price for the rate of exchange between two assets. In the context of Stablecoins, the peg refers to the specific price that a token is aiming to stay at – aka the target value a Stablecoin seeks to maintain over a long period of time – this is the “Stable” in Stablecoin! 😊 . The vast majority of Stablecoins are pegged to one US Dollar – aka each coin is worth $1.

Collateralized Stablecoins maintain their peg through contraction (removing or “burning” existing coins) or dilution (adding or “minting” new coins) to affect the total supply. The changes in the coin supply will change the relative price of each coin, until it reaches the desired peg. Algorithmic Stablecoins maintain their peg through a combination of collateralization and utilizing complex block chain based smart contract algorithms that contract and expand the supply based on various market factors.

Depegging

A depegging event occurs when a stablecoin loses its intended 1:1 parity with its underlying asset –and drops in price well below $1. Depegging events are hugely important to avoid first and foremost because it serves to destabilize the coin which negatively affects trust and market confidence in the crypto ecosystem and brings financial risk directly to the buyers/users of Stablecoins.

Holders of a depegged coin face the possible immediate loss of capital as they may not be able to redeem coins for the fiat currency backing the coin because buyers flee the market – aka a bank run scenario builds rapidly as users rush to redeem coins simultaneously.

Stablecoin protocols may be forced to register their collateral asset positions as undercollateralized, triggering automated liquidations that force asset sales which further accelerate worsening price declines.

Contagion effects can develop where one stablecoin’s depegging can destabilize other stablecoins and even negatively impact cryptocurrencies like bitcoin. Lastly, the Stablecoin’s issuers reputation can be severely damaged – potentially forever – possibly leading to regulatory crackdowns.

Uses Cases – What are Stablecoins good for?

Once used primarily for crypto trading, particularly smoothing the on and off ramp for purchases of Bitcoin and /or Ethereum for example, Stablecoins have now become a versatile tool for multiple uses.

Stablecoins are becoming the backbone for facilitating lending, borrowing, and yield farming. Their lack of price fluctuation makes them ideal for liquidity pools, where they reduce impermanent loss and maintain the efficiency of decentralized exchanges (DEXs). Stablecoins also enable global access to financial services, empowering users in economically unstable regions to participate in foreign financial markets without exposure to local currency volatility.

Their ability to process transactions quickly and cost-effectively, 24x7x365, often with minimal fees compared to traditional banking systems, makes them an attractive option for peer-to-per (P2P) transactions by providing a simple and secure way for individuals to exchange value without intermediaries – particularly for cross-border payments and remittances. Stablecoins simplify transactions for importers and exporters, providing a stable and transparent medium for international trade, particularly in regions with limited access to foreign currency.

U.S. Regulation – SEC and the GENIUS Act

SEC’s position on Stablecoins

The role(s) of the U.S. Security and Exchange Commission (SEC) with respect to Stablecoins has been under debate (rather boisterous at times) for the past 4-5 years (2021-2025). The core question is whether the SEC should regulate Stablecoins as securities or whether they should fall under banking/payment regulations instead.

On April 4, 2025, the SEC’s Division of Corporation Finance published a “Statement on Stablecoins” in which it states that certain Stablecoins – aka “Covered Stablecoins”, are not securities under federal law, so their issuers don’t need to register them with the SEC.

In the statement “Covered Stable” coins are defined as:

• Pegged 1:1 to USD value - designed to maintain stable value equal to one U.S. dollar per stablecoin

• Redeemable 1:1 – users can exchange any stablecoin back for exactly $1 USD at any time, with no limits

• Having a Fully Backed Reserve – Stablecoins are backed with a reserve of low-risk, highly liquid assets (like cash or U.S. Treasuries) worth at least as much as all Stablecoins in circulation

The SEC’s rationale for this position goes something like this …

Users acquire Stablecoins to use them for a number or purposes such as making payments, transmitting money, or storing value—like a “digital dollar.” As such, users are not viewing Stablecoins as investments, and are not expecting profits, interest, or ownership rights – but users do want stability.

Users are not expecting share or profit from the issuer’s financial performance. Because Stablecoins are fully backed with safe, secure, liquid assets leveraging securities laws is not necessary. Net-Net, the SEC currently views that Banking Regulators (Treasury OCC, State Bank Regulators) are better suited addressing “near-money” assets like Stablecoins.

In July 2025, the GENIUS Act passed, assigning payment stablecoin regulation to Treasury OCC (federal) and state banking regulators, not the SEC.

Illicit Activity with Stablecoins – The Dark Side Exists

High-risk and illicit actors are attempting to leverage Stablecoins for a variety of illegal activities. Their stability and global accessibility make them attractive tools for bad actors seeking to bypass financial controls and avoid detection — although the inherent transparency and traceability of blockchain often makes this a poor choice.

Stablecoins have been used in money laundering, fraud, and sanctions evasion. Due to their relatively high liquidity and acceptance across cryptocurrency exchanges, Stablecoins can be used to transfer value quickly across borders without relying on traditional financial institutions.

Sanctions evasion through Stablecoins and other cryptocurrencies has gained prominence as countries like Russia and Iran explore alternatives to bypass Western financial restrictions often through complex networks of wallets and exchanges.

Stablecoin issuers have stepped up their efforts to fight financial crime, supporting global law enforcement and regulatory investigations. Most centralized stablecoin issuers have the power to freeze or permanently delete or “burn” tokens in wallets linked to confirmed criminal activities, to comply with regulations, and help to stop illegal transactions and helping recover stolen funds.

Issuers like Tether work closely with global law enforcement agencies, financial crime units, and regulators like the Financial Crimes Enforcement Network (FinCEN) using Chainalysis to monitor transactions in real-time and identify suspicious activity.

Chainalysis plays a critical role in proactively assisting in the detection and prevention of illicit activities involving Stablecoins. With real-time monitoring, Chainalysis can identify frozen or burned assets, trace the flow of funds, and map networks of wallets associated with high-risk actors across various blockchains, reinforcing trust in the ecosystem and allowing innovators to confidently build onchain.

Adoption – Where are Stablecoins Used

Latin America and Sub-Saharan Africa are the fastest growing regions for retail and professional-sized Stablecoin transfers, with year-over-year (YoY) growth exceeding 40%. Eastern Asia and Eastern Europe follow closely, with 32% and 29% YoY growth respectively.

Across the Middle East and North Africa, Stablecoins and altcoins are capturing a larger share of the market, particularly in Turkey, Saudi Arabia, and the UAE.

In Eastern Asia, Hong Kong is seeing a flurry of interest from potential issuers. In Central and Southern Asia and Oceania, Stablecoins are widely used for cross-border trade and remittances, bypassing traditional banking challenges. Countries like Singapore have bolstered Stablecoin confidence through regulatory frameworks, making Stablecoins a key tool for both retail and institutional users.

Meanwhile, markets like North America and Western Europe have seen meaningful, but slower growth rates of retail Stablecoin activity, likely due to robust native financial infrastructure, and the existing relatively high levels of Stablecoin usage. In Western markets the next wave of Stablecoin adoption will be institutional investors (Banks, Investment organizations) increasingly adopting Stablecoins for liquidity management, settlements, and entry into cryptocurrency.

Are Stablecoins Safe?

If it has not dawned on you yet, Stablecoins might seem like a low-risk investment compared to volatile cryptocurrencies – and that is partially true. No Stablecoin is risk-free and trust depends entirely on the issuer’s credibility and reserve asset quality

Further, Stablecoin safety depends heavily on their design, backing and the regulatory environment they operate in. Regulations across the globe vary wildly. Stablecoin issuers may refuse or fail to redeem tokens at face value, which can be the single biggest risk to holders.

Some Stablecoins are safer than others. Fiat-backed Stablecoins such as USDT and USDC are considered safer than some other Stablecoins because they’re backed by reserves of cash or government bonds. Crypto-backed Stablecoins, such as DAI, maintain their dollar peg by using over-collateralized cryptocurrencies locked in smart contracts, making them vulnerable to the volatility of the underlying crypto assets and the potential for technical flaws in their supporting smart contracts that could expose the Stablecoin to hacking.

Among the riskiest types of Stablecoins are Algorithmic Stablecoins which rely on market incentives and algorithms to maintain their value. With Algorithmic coins you are placing your trust in the robustness, stability, security, and ability of the algorithmic protocols to deal with rapidly accelerating price declines, or other unforeseen market forces, and the potential for depegging events.

Net-net, the current consensus is that Stablecoins are comparatively safe within the crypto world but do carry real financial risks. Users are encouraged to treat Stablecoins as high-risk financial instruments, and to NOT view Stablecoins as FDIC-insured bank deposits – which they are absolutely not. For maximum safety, pundits will suggest using only USDT, USDC, or other top-rated, fully audited, regulated fiat-backed stablecoins—and never hold more in Stablecoin(s) than you can afford to lose if the issuer fails.

Wrapping up

Stablecoins like all crypto powered products have inherent risks .That said, there is growing consensus that Stablecoins are rapidly becoming a core component of the global financial system, and yes, it is safe to say that this area of the financial industry will continue to evolve rapidly and adapt to address new use cases, applications, and opportunities. Digital Dollars are here to stay!

Please, share your thoughts via the Comments section for this post or open a new SubStack chat thread … and please forward this post to your friends and colleagues. Until next time!

Let’s Get Started!

SpaceX’s Starship is a two-stage, fully reusable heavy lift rocket launch system made up of the Super Heavy Booster lower stage and the Starship upper stage. Both stages are powered by the latest generation Raptor V3 engines that burn liquid methane (CH4) and liquid oxygen (LOX).

The lower Super Heavy Booster uses 33 Raptor engines arranged with 13 center engines and 20 outer engines. The Starship upper stage uses six engines - three Sea-Level Raptors and three Raptor Vacuum engines (RVacs) for operation in space. (See engine details below for additional specs). The bodies of both stages are made from stainless steel with domes inside the stages to form tanks for the methane and oxygen fuel. Both stages are intended to re-enter the earth’s atmosphere and land vertically at the launch site post mission – potentially to re-used in additional missions.

Space X positions Starship as a space transportation system designed for Earth orbit, travel to the Moon and beyond (Mars), and point-to-point travel on Earth. The full rocket stack is about 123 Meters tall, 9 Meters in diameter, with a fully reusable payload capability of 100–150 metric tonnes and an expendable capability up to 250 metric tonnes. Starship is designed to be the successor to SpaceX’s Falcon 9 and Falcon Heavy launch vehicles.

Overview of Starship’s Origins and Design Principles

Space X entered the commercial space industry with the Falcon 1 rocket, a two-stage liquid-fueled launch vehicle designed to send small satellites into Low Earth Orbit (LEO). Falcon 1 was much cheaper to build and operate than its competitors, made possible largely by the SpaceX-developed Merlin engine combined with SpaceX’s focus on making rockets reuseable.

In March 2006, SpaceX attempted its first Falcon 1 launch, which began successfully but failed in flight due to a fuel leak and fire. Launches in March 2007, and August 2008 failed to reach earth orbit. The launch in September 2008 was successful, making SpaceX the first privately owned company to send a liquid-fueled rocket into orbit.

In 2010 SpaceX launched Falcon 9, a much larger vehicle powered by 9 engines. Falcon 9 was designed so that its first stage could be reused, and in 2015 a Falcon 9’s first stage successfully returned to Earth near its launch site. In 2010, SpaceX began developing the Falcon Heavy vehicle with its first test flight occurring in 2018. Two of the first three stages landed successfully; with the third hitting the water near the landing drone ship. This test flight did not carry a satellite but instead placed a Tesla Roadster with a mannequin in a space suit buckled into the driver’s seat into orbit around the Sun. The first operational flight of the Falcon Heavy launched on April 11, 2019.

The Starship heavy lift vehicle evolved from a sequence of much larger travel to Mars vehicle concepts inside SpaceX, that were ultimately narrowed into the current two-stage Starship/Super Heavy system.

Starship is not a direct descendant of Falon Heavy but rather was developed in parallel with the vehicle built from Falcon 9 cores. Starship became a ship-plus-booster architecture rather than a single giant launch vehicle, with orbital refueling as a central requirement for deep-space missions. (See refueling section below for details) Ultimately, Starship is intended to supersede the whole Falcon-era family with a fully reusable architecture built for interplanetary transport as SpaceX’s next-generation heavy lift space launch vehicle.

Starship’s Connection to NASA’s Artemis Moon Landing Program

Formally established in 2017, NASA’s Artemis program is a Moon exploration program that aims to land humans on the Moon for the first time since Apollo 17 in Dec 1972. It also intends to establish a permanent base on the Moon in the 2030s.

NASA picked SpaceX to build one of two lunar landers for Artemis with Blue Origin contracted to supply a second lunar lander. According to NASA, the next Artemis mission, Artemis III, is currently scheduled for “late 2027”. The plan for Artemis III has NASA launching the Orion crew capsule on NASA’s SLS rocket. Orion will then dock with the Starship-based lunar lander in low earth orbit. The mission will focus on executing docking tests with the SpaceX lunar lander, and possibly the Blue Origin lunar lander while in earth orbit.

The first lunar landing in the Artemis program is planned for the Artemis IV mission, sometime in 2028. In Artemis IV, the Starship based lunar lander aka “Human Landing System” (HLS) is to be launched into a low Earth orbit and then refueled by multiple Starship tanker spacecraft. Once fueled, it would perform a trans lunar injection burn, travel to the moon, and enter a near-rectilinear orbit around the moon.

Next the Orion spacecraft would be launched via NASA’s SLS rocket, travel to the moon, and dock with the waiting HLS lander. After two to four of the crew have transferred into the HLS, it would depart orbit and descend to the lunar surface. After completing lunar surface operations, the HLS will lift off from the Moon and return to lunar orbit to rendezvous with Orion. The crew then transfers back to Orion and departs for Earth.

While Starship will not transport astronauts to and from the moon in Artemis IV, it will likely have role to play in the future of the Artemis program. No details have been announced about Starship’s possible role(s) in the future.

Late Breaking News!

SpaceX: May 27, 2026, FAA Orders SpaceX to “conduct a mishap investigation” to investigate and report on the failure of RaptorV3 booster engines that occurred during the May 22, 2026 Starship test flight #12.

After separating from the upper stage of the rocket, the Super Heavy Booster engines failed to reignite and execute the boostback burn resulting in the booster tumbling back to earth and most likely exploding on impact.

This means SpaceX will have to pause any further Starship test launches until the investigation is completed.

____________________________________________________________________________

Blue Origin: May29, 2026, during a “HotFire” test of Blue Origin’s New Glenn rocket, the rocket exploding seconds after the start of the test, engulfing and destroying the launchpad in a giant fireball.

Starship Stages and Core Components

Starship - Upper Stage

Starship, the upper stage of the rocket, contains 4 primary sections – the engine bay, the oxygen tank, the fuel tank, and the payload bay. Starship is powered by six Raptor engines, which are housed within a dedicated shielding compartment. Engines include 3 optimized for atmospheric operation (sea-level engines), as well as 3 engines optimized for operation in the vacuum of space, called RVacs. Each engine is protected by a dedicated shielding compartment.

Starship upper stage – is constructed with 3.97 mm (0.156 in) thick stainless steel cylinders. The stage measures 52.1 meters (171 ft) tall, 9 meters (30 ft) wide and weighs (dry, no fuel) ~ 100 tonnes (220,000 lb). Upper stage fuel tanks can hold 1,200 to 1,500 metric tons of propellant.

Domes inside the stage separate the methane and oxygen tanks. SpaceX has stated that Starship, in its “baseline reusable design”, will have the capability to get a payload capacity of 100–150 tonnes (220,000–331,000 lb) into low Earth orbit and 27 tonnes (60,000 lb) into geostationary transfer orbit.

The windward side is protected by a heat shield, which is composed of eighteen thousand hexagonal black tiles that can withstand temperatures of 1,400 °C (2,600 °F), and are designed to be used multiple times with minimal maintenance between flights.

Fuel Tanks

The propellant tanks in the upper stage are separated by a common bulkhead, similar to the ones used on the Saturn V rocket. The upper stage tanks hold 1,500 tonnes (3,300,000 lbs) of propellant, consisting of 1,170 tonnes (2,580,000 lbs) of liquid oxygen and 330 tonnes (730,000 lbs) of liquid methane.

Fuel is fed to the engines via four downcomers, with three smaller downcomers feeding the RVacs and the central downcomer feeding the inner three engines. The central downcomer connects to a large sump, instead of directly to the methane tank itself.

The oxygen tank terminates with the thrust structure of the vehicle. The RVacs are mounted directly to the aft dome, which has reinforcements mounted inside of the tank. The three sea level engines are mounted on the thrust puck, which forms the bottom of the aft dome. A conical steel structure is mounted inside the bottom of the dome, reinforcing the thrust puck enough to enable its support of the inner three engines. The propellant lines on the vehicles are all vacuum jacketed, reducing boiloff while in orbit.

During unpowered flight in orbit, control authority is provided by cold gas thrusters fed with residual ullage gas. Four such thrusters are located just below the payload bay, and two on the oxygen tank.

Once in space, the Starship upper stage is intended to function as a standalone spacecraft capable of carrying crew and cargo.

Super Heavy (lower booster) Stage

Super Heavy uses 33 Raptor engines fueled by sub-cooled liquid methane (CH4) and liquid oxygen (LOX), arranged with 13 center engines and 20 outer engines. The outer 20 engines, arranged in a ring, are fixed in place. To save weight, the 20 engines are started using ground support equipment on the launch mount and cannot be reignited for subsequent burns. The inner thirteen engines are equipped with gimbal actuators and are designed to reignite for the boostback burn to support reentry and landing burns.

Super Heavy – is 72.3 meters (237ft) tall, and 9 meters (30 ft) wide, and contains four general sections: Raptor engines, the liquid oxygen tank, the liquid methane tank, and the interstage. This lower stage weighs ~275,000 kg (606,000 lbs) (dry, no fuel), ~3,675,000 kg (8,106,000 lb) when loaded with fuel.

Booster Fuel Tanks

Similar to the upper stage, the booster stage has two cryogenic propellant tanks separated by a common bulkhead, a similar the Saturn V rocket. Each tank possesses roughly 74 stringers for structural reinforcement, attached to their interior walls.

The booster’s two tanks hold a combined 3,400 tonnes (7,500,000 lbs) of propellant, 2,700 tonnes (6,000,000 lbs) of liquid oxygen and 700 tonnes (1,500,000 lbs) of liquid methane. Fuel is fed to the engines via a single liquid downcomer and channeled into distribution manifolds of the engines.

The oxygen tank ends at the thrust puck of the vehicle. While the outer twenty engines are mounted on a ring attached to the first steel ring, the inner thirteen are mounted onto the thrust puck, a part of the aft dome. Large steel structures are attached to the bottom of the dome, reinforcing the puck sufficiently to fully support the inner thirteen engines, and at the same time providing pathways for methane and oxygen into the engines.

Interstage

The interstage is the structural section connecting the Super Heavy booster (lower stage) to the Starship spacecraft (upper stage). Its primary purpose is to enable hot staging—a separation method where the upper stage’s Raptor engines ignite before the booster fully shuts down—while also providing critical structural and operational functions.

The interstage is equipped with four electrically actuated grid fins made of stainless steel, each with a mass of roughly 3 tonnes (6,600 lb). The fins remain extended during ascent to save weight, though this results in mild warping during stage separation. The grid fins are used to orient the decent flight of the booster to be captured by the launch tower.

The interstage also has protruding hardpoints, located between grid fins, allowing the booster to be lifted or caught by the launch tower, nicknamed “Mechazilla”. The ability to lift a booster from these hardpoints was proven on August 23, 2022, when a Booster was successfully lifted onto the “Orbital Launch Mount” A (OLM A). The first live catch of a booster successfully occurred on October 13, 2024.

Payload Bay

The payload bay hosts the nosecone, header tanks, forward flaps, multiple Composite Overwrapped Pressure Vessels (COPVs), and the satellite “PEZ dispenser”. COPVs are high-pressure tanks made of a thin metal liner overwrapped with strong composite fibers (like carbon fiber) to save weight while holding pressurized gases. In Starship, COPVs contain gaseous nitrogen used for pressurization systems. The header tanks are mounted at the tip of the payload bay. Several COPVs are mounted in the space around the methane header tank, providing the startup gas for the engines, with twelve additional COPVs within the base of the payload bay.

The LOX header tank forms the top of the nosecone, with the methane header tank attached directly below it. These tanks terminate in a conical sump, which are attached to the downcomers.

The nosecone has substantial internal reinforcement via internal stringers, mainly around the forward flap attachment points and lifting points for the capture chopsticks used to catch the rocket.

The PEZ dispenser is used to deploy Starlink satellites into orbit. It consists of the dispenser mechanism and the door, with the door opening by folding into the payload bay. A mobile track is used in the base, enabling the dispenser to push the satellite out of the vehicle. After releasing a satellite, the next payload is lowered onto the base and is released. The opposite occurs during loading, with the dispenser raising its payloads to support receiving and loading another satellite.

Re-Entry Flaps

Starship controls its reentry capture with four flaps; two aft flaps mounted to the sides of the engine bay and LOX tank and two forward flaps on the payload bay. Substantial reinforcements are present in the nosecone for the support of the forward flaps. SpaceX indicate the flaps replace the need for wings or tailplane, reduce the fuel needed for landing, and allow landing at destinations where runways do not exist (for example, Mars).

StarShip Configurations

SpaceX initially proposed using Starship for crewed missions to Mars, though SpaceX has not published technical plans or designs about Starship’s life support systems, radiation protection, docking system, or in-orbit refueling system for Mars.

This plan was later revised to prioritize on the shorter term objective of returning humans to the Moon. In addition to supporting returning humans to the moon and/or building a permanent base on the moon, SpaceX has proposed a wide range of missions for Starship, such as deploying large satellites, space station modules, space telescopes, and executing point-to-point cargo transport flights on Earth.

Starlink

SpaceX plans to use Starship to launch the second generation of satellites for its Starlink system, providing higher volume of satellite deployments into orbit, making Starlink launches cheaper.

Transportation

SpaceX has proposed using Starship for point-to-point flights - called “Earth to Earth” - traveling anywhere on Earth in under an hour. SpaceX indicates that it would complete hundreds of cargo flights before launching with human passengers.

Raptor 3 Engines

The engine in Starship is the Raptor 3, the third generation of the engine, a reusable methane-oxygen staged-combustion engine. The Raptor was originally conceived to burn hydrogen and oxygen propellants as early as 2009 and was intended to be several times more powerful than the Merlin series of engines. In November 2012, SpaceX disclosed it was working on methane-fueled rocket engines, and that Raptor would be methane-based.

Raptor 1 proved the concept, Raptor 2 made the engine lighter and able to run reliably at scale, and Raptor 3 is the production-optimized version intended to power Starship going forward. Through 2014 -2021 development and testing of Raptor 1 and 2 proceeded with Raptor 2 entering production in late 2021. In 2022, Raptor 2 achieved consistent operation with 11 test flights powered by Raptor 2 executed in 2023 through 2025. On May 22, 2026, Starship test flight 12 was the first SpaceX flight powered by the Raptor 3 engine.

Raptor 1 proved that full-flow staged combustion with methalox (liquid methane + liquid oxygen) could work in practice and it successfully supported static testing. Raptor 2 was the first mass production version of the engine and delivered multiple successful full-stack test flights that demonstrated booster-stage performance. Raptor 3 delivers multiple refinements including increased thrust, lower weight, and a reduction in the number of parts which improves re-use economics.

Staged Combustion Cycle

Raptor uses a full-flow staged combustion cycle where propellant flows through multiple combustion chambers and is thus combusted in stages. Both propellants are partially burned in separate preburners, where a small portion of propellant is partly combusted and the preburner exhaust streams drive turbopumps that inject the hot gas into the main combustion chamber. This design is more efficient in burning propellant to create thrust but involves more complex engineering.

In full-flow combustion both propellants are fully gasified before final combustion, mixing is faster and combustion is more complete. The turbine exhaust is cooler than in many other staged-combustion engines, which helps engine durability and reuse. SpaceX also benefits from very high chamber pressure, which improves performance but increases design difficulty.

A useful mental model is that Raptor is not just “fuel plus oxidizer burns in a chamber”, rather it is a highly integrated pump-and-preburner system where both propellants are conditioned into hot gas first, then recombined in the chamber for maximum performance.

Propellant Flow Path and Combustion Details:

The propellant flow path sequence is as follows:

• Liquid methane and liquid oxygen are fed from the vehicle tanks into separate pump systems.

• Each propellant passes through its own preburner, where a small fraction is burned to create hot, high-pressure gas.

• The methane-rich gas drives the fuel turbopump, and the oxygen-rich gas drives the oxidizer turbopump.

• After driving the turbines, both streams are injected into the main combustion chamber, where methane and oxygen are mixed and burned for thrust.

SpaceX chose methane because it burns cleaner than kerosene (used in the Merlin series engine), which reduces carbon buildup and helps with reusability. The full-flow architecture also supports long-life turbomachinery because neither turbine must tolerate the full thermal stress of a single rich preburner stream. Compared with Merlin’s open-cycle liquid kerosene (RP-1) + oxygen design, Raptor is much more complex but improves reusability.

Liquid Methane (CH4) – the major constituent of natural gas, is the primary fuel used in Starship/Raptor. Compared to other hydrocarbon fuels, methane produces less carbon dioxide for each unit of heat released. As rocket fuel it produces smaller exhaust molecules compared to other fuels, which reduces the deposition of soot on engine components. Also, methane is easier to store than hydrogen due to its higher boiling point and density. Methane is handled at roughly-256°F (−162°C) and is further subcooled for launch loading.

Liquid Oxygen LOX – serves as the fuel oxidizer in Starship and is cryogenic liquefied at around -297°F (-183°C) then subcooled further for densification. LOX is non-toxic but reactive, making it a preferred, cost-effective oxidizer for large propulsion systems despite its cryogenic handling requirements.

SpaceX uses liquid nitrogen-based pad infrastructure and subcoolers to prep propellants before loading into Starship.

Engine Variants

The Raptor engine has sea-level and vacuum-optimized (RVacs) variants. Starship’s upper stage uses vacuum engines for orbital efficiency, while Super Heavy booster uses sea-level Raptors for liftoff and boostback burns. The vacuum version uses a larger nozzle expansion ratio to raise exhaust velocity in low pressure environments (in space), that lets more exhaust energy turn into useful forward motion.

The sea-level engine trades some vacuum efficiency for better behavior during launch, ascent, and landing in the atmosphere, so the sea-level version is better suited for booster use.

Also, the sea-level engines are gimbaled for thrust vector control, while the Raptor vacuum engines are fixed and do not gimbal. In the Raptor engine, the nozzle itself is not a separate gimbaled nozzle, rather the engine/nozzle assembly pivots to steer thrust which is used to control pitch and yaw during flight.

Orbital Propellant Transfer – The Artist Known as In Orbit Re-Fueling!

Starship will implement orbital propellant transfer—a set of highly orchestrated procedures for launching specialized re-fueling tanker Starships into Low-Earth Orbit (LEO) that then dock with a mission Starship and transfer liquid methane (CH₄) and liquid oxygen (LOX) to refuel the mission spacecraft for deep-space missions – aka for traveling to the moon or elsewhere!

Tanker spacecraft are essentially standard Starships without windows, optimized to carry the maximum volume of propellant. SpaceX estimates it will take approximately ten tanker launches of propellant to a depot in orbit to refuel a Starship sufficiently for it reach orbit around the moon. SpaceX successfully conducted a test in March of 2024, where cryogenic propellant was transferred between two tanks inside a space craft.

Propellant flow is achieved via a pressure differential between the donor tank and the receiving tank on the two spacecraft. This is a simpler solution than relying on pumps, but if the pressure differential method comes up short, SpaceX may need to consider employing pumps, making this procedure much, much more complicated.

Starship docking is an autonomous process that relies on LIDAR sensors, developed and used for SpaceX’s Dragon spacecraft. The sensors determine distances and support alignment calibrations, while machine vision tracks the docking target.

The two spacecraft will dock nose-to-nose, not side-to-side, and fluid transfer connections will use the same ports used to load propellants into the rocket on the launch pad. SpaceX will fine-tune tank pressures and fire propellant settling thrusters to enable fluid flows. The slow boil-off of cryogenic propellant in the Starship refueling tanker should raise pressures in the donor tank to help drive methane or liquid oxygen into the receiving tank. SpaceX can also adjust pressures using header tanks at the top of each ship.

Step by Step Re-Fueling Procedures looks like this:

• First, a target Starship launches to LEO

• The target ship is kept “alive” in orbit with extra power and batteries at last long enough for the second ship to arrive

• 3–4 weeks later, a tanker Starship launches and docks with the target Starship

• Docking is nose-to-nose at the nose/cone section, avoiding heat shield contact and uses quick-disconnect (QD) plate ports for propellant lines.

• Once docking is locked, umbilical lines connecting the propellant systems of both vehicles are engaged, creating the physical connection

• Fluid transfer via pressure differential between tanks (donor higher, receiver lower) will move cryogenic propellant into the receiving tanks without pumps

• Settling thrust; cold-gas (nitrogen) thrusters provide milli-g acceleration (0.001G) to keep liquid propellant near the transfer values at tank outlet, preventing gas bubbles from entering transfer line

• When transfer is complete, tanker undocks, re-enters atmosphere, and is caught by the launch tower, target Starship remains in orbit as propellant depot

When the tanker Starship loaded with propellant comes in for docking, the ship’s maneuvers to approach its counterpart could induce sloshing of fluid inside the tanks. This could introduce unforeseen forces that the ship’s navigation system must account for during the rendezvous and docking procedure.

NASA has flagged that SpaceX must understand the slosh dynamics of propellants in the tanks as Starship maneuvers, as well as the amount of settling thrust needed once the vehicles are docked to ensure propellant flows between them.

Also, in microgravity, liquid propellants don’t stay conveniently pooled at the bottom of their tanks — they float freely. If a spacecraft applies thrust, the propellant will stay still until it splashes against its tank walls. If a spacecraft thrusts in one direction and opens a valve on the tank in the opposite direction of that thrust, the propellant — attempting to stay at rest — will naturally escape out of that opening. Thus, if a spacecraft in need of fuel docks with a tanker, their tanks are connected, and the tanker attempts to accelerate away from the receiving ship, the propellant in the tanker’s tanks will effectively be pushed into the second ship as the fluid tries to stay at rest.

To address the complications of microgravity, the system uses cold-gas settling thrusters that create a slight artificial gravity of around 0.001G, which helps keep the liquid propellants near the transfer valves for a stable flow during the process. This is the “settling thrust” step listed in the step-by-step re-fueling process described above.

Storage of cryogenic propellant long-term while in LEO is difficult because even small amounts of heating steadily warm the tanks, causing boil-off, pressure rise, and propellant loss over time. For Starship methane/oxygen systems, the temperatures used are less extreme than liquid hydrogen, but the same basic problem still applies: the vehicle must keep propellant cold, stable, and usable for days to weeks in a warm (relative) orbital environment

Full-Scale Demonstration

For the first full-scale refueling demonstration, is planned for “late 2026”. First, SpaceX will launch a Starship to serve as a target vehicle in low-Earth orbit. This ship will have an augmented power system and more battery capacity to sustain itself in space long enough for the launch of the second vehicle — the refueling tanker. Both ships involved in the refueling demo will have thermal insulation and vacuum jacketing on propellant lines and around internal plumbing to minimize boil-off.

As SpaceX continues Starship test flights, Engineers will be assessing and measuring the boil-off rates of methane and liquid oxygen in space. It is critical to understand how much propellant is lost from boil-off on order to know how many refueling tankers they need to launch for a Starship lunar landing mission.

Implications for Launch Pads and Earth based Support Systems

To support orbital re-fueling and other moon mission programs, SpaceX will need to ramp up the Starship launch cadence of many flights from multiple launch sites, and concurrently SpaceX will need to capture, and rapidly refurbish and reuse Starships and Super Heavy boosters. This will likely require building and operationalizing additional launch pads, tower and support facilities including ground based cryogenic fuel systems both in Texas and in Florida.

Potential for Fuel Supply Bottlenecks

As SpaceX seeks to accelerate the pace of Starship launces it faces a potential bottleneck with getting enough liquid methane, liquid oxygen, and liquid nitrogen to each launch site in a way that supports a much higher launch cadence. Note that a single Starship launch can require more than 200 tanker-truck deliveries of methane, oxygen, and nitrogen, which is slow, expensive, noisy, and polluting, and has the potential to become a serious constraint if SpaceX wants to achieve anything close to the launch frequencies it has discussed for Texas and Florida.

SpaceX has not published a definitive fixed annual launch rate for Starship, but its public goal is to support a “very high cadence” of possibly getting to dozens of launches per year and eventually much more. In the near term, the clearest public benchmark is that Starship is designed around rapid reuse and large-scale orbital operations, which is why SpaceX is building out its launch-site propellant infrastructure. The likely solution set for increasing fuel supplies is a mix of larger on-site storage, on-site cryogenic production, and eventually pipelines or other dedicated logistics so launches are not constrained by tanker-truck deliveries.

Wrapping up

For NASA’s Artemis III and IV missions to be successful, there is a TON of spacecraft design, engineering and orbital operational work to be accomplished. Including work to refine and improve the Starship launch vehicle. SpaceX Starship is one of the core components that must function perfectly for the next series of Artemis missions to accomplish their goals. Watch this space for future developments and stay tuned multiple heavy lift rocket launches!

Please, share your thoughts via the Comments section for this post or open a new SubStack chat thread … and please forward this post to your friends and colleagues. Until next time!

Let’s get started!

Like the global “race” for AI leadership, Quantum Technology has enormous geo-political dimensions that span multiple domains including national security, technological sovereignty, economic advantage/disadvantage, and military power.

Quantum Technologies are clearly viewed as increasingly strategic assets, as critical as energy infrastructure (Oil, Gas Electric Grid) and/or nuclear technologies. The siren calls now heard in governments around the world is “We cannot afford to fall behind” in Quantum and there are natural comparisons being made between today’s Quantum race and previous nuclear and space races.

A core underlying dilemma is not whether to collaborate or compete, but how to do both without weakening security, losing control over IP, or slowing innovation. One way to think about this question is this: Quantum is still early enough that countries need each other to build the technology, refine standards and implement supply chains – but it is strategically important enough that countries and commercial enterprises are already trying to outpace each other and develop strategic advantage so they can control Quantum for their benefit.

Why Competition rises and why Collaboration still Matters

Quantum technologies are widely viewed as dual-use, meaning the same advances can support civilian innovation and military advantage, which has the direct potential for intensifying rivalries among nation states. This makes governments more likely to subsidize domestic champions, limit sensitive transfers, and treat Quantum progress as a “race” rather than a shared scientific project. Competition is also amplified by the possibility that early leadership could shape standards, supply chains, and future market power.

No single actor can easily master the full stack of Quantum technology alone, so collaboration remains essential for research, bench marking, knowledge building, workforce development, and scaling production. Alliances can help share talent, coordinate export rules, build interoperable supply chains, and spread the cost and risk of long-term R&D. In that sense, collaboration is not the opposite of competition; it is often the mechanism that makes competition possible.

The phrase “open enough to move the field forward, closed enough to protect strategic advantage” appears to be the most common operational pattern we see nation states and commercial enterprises using today.

Government investment in Quantum technologies has accelerated over the past 5 years and is currently estimated to be north of $50B worldwide. A few examples, and bear in mind these tend to be best guess estimates and do not include private sector and/or venture capital investments:

• China has dedicated more than $15B across multiple Quantum projects, including $10 billion for its National Laboratory for Quantum Information Sciences.

• The EU has targeted ~ $11-12B towards Quantum technologies.

• The U.S.A. government has committed ~$5-8B, with $1.8B allocated via the National Quantum Initiative Act.

• The UK has pledged ~ $3B via its 2023-2033 National Quantum Strategy.

Naturally much of the Geo-Political commentary focuses on the U.S.A. – China rivalry, and not surprisingly the EU has also initiated and funded multiple programs and industry consortium to protect Europe’s interests and to blunt potential reliance on U.S. and Chinese innovation. Another 18 countries including Australia, Canada, Denmark, France, Germany, India, Israel, Japan, Norway, Russia, Singapore, and Spain (to name some) are accelerating their investments into Quantum technologies.

Not surprisingly, China is seeking independence and control and is developing its own Quantum hardware, software, and even cryptographic standards so to avoid reliance on western technology. Even middle eastern and gulf countries including the United Arab Emirates, Saudi Arabia, and Qatar view Quantum as an opportunity to extend and diversify their economies beyond oil and gas and are investing in Quantum technologies. NATO views Quantum technologies as a key element of strategic global competition, with significant potential for both the alliance and its adversaries.

National Security, Cybersecurity, and Cryptography

It is well documented that Quantum computing will at some point in the future easily defeat existing encryption algorithms and cryptographic systems and the point in time when this will be possible has its own label > aka “Q-Day”.

Modern encryption protocols (securing everything from internet access, personal and corporate data, communication systems, military secrets, cryptocurrencies to online banking) rely on mathematical problems so hard to compute that classical computers can’t crack them in a reasonable amount of time.

See this post on how ECDSA Cryptography’s Role in Securing Cryptocurrencies.

Q-Day is the point in time when a cryptographically relevant Quantum computer (CRQC) becomes capable of breaking today’s widely used public-key encryption algorithms like RSA and Elliptic Curve Cryptography (ECC). Q-Day is not a fixed date but a capability threshold with some estimates suggesting it could happen as early as 2030-2033.

National Security

Quantum computing when fully functional could break today’s core encryption standards, endangering the security of ALL communications, critical infrastructures, and ALL digital data systems worldwide. The implications for national security are global in scale and massive in scope and pose the real possibility of amplifying existing geopolitical tensions and heightened needs for strengthening defensive alliances. For example, NATO views Quantum technologies as a key element of strategic competition, with significant potential for both the Alliance and its adversaries.

Global powers, led by China, the U.S.A., and the EU are racing to build Quantum supremacy. A country that develops Quantum systems first could gain the ability to decrypt rivals’ secret communications, giving it profound intelligence, economic, and military advantages. Some analysts are describing a rapidly evolving Quantum arms race, with major powers working to upgrade their cybersecurity processes, systems and technologies with Post Quantum Cryptography – aka the artist known as PQC.

Further, National Security implications are dual mode as there is the need to balance offensive application of Quantum computing, such as cracking encryption used by adversaries, with building strong defensive postures – aka securing your own assets and digital systems from Quantum based attacks.

Proactive transitions in security and cryptographic systems and technologies – Post Quantum Cryptography - PQC

All organizations, public and private should have begun preparing to transition digital systems and communications for post-quantum encryption by implementing protections prior to Q-Day or risk having their entire digital infrastructure compromised.

At the national level, governments need to prioritize the regulatory requirements for the security of mission-critical operations and develop new Quantum resistant standards for critical infrastructure. At a global level, cooperation around this transition is also important given the interdependency of digital infrastructures.

The good news is that the preparatory work has begun – here are two key examples:

NIST Cybersecurity Standards – PQC

The next evolution in cryptography is referred to as Post-Quantum Cryptography (PQC) and is being developed to counter the threat of Quantum computing being able to crack existing cryptographic systems. Simply stated, PQC is developing new cryptography methods that can withstand Quantum attacks with the goal to prevent “Q-Day” causing harm to digital infrastructure.

In the U.S., the National Institute of Standards and Technology (NIST) recognized the need to standardize Quantum-secure cryptographic primitives and On August 13, 2024, NIST released final versions of its first three Post Quantum Cryptography Standards. The standards provide detailed descriptions of post-Quantum encryption and digital signature algorithms so they can be implemented consistently to facilitate secure and interoperable communication.

These new standards are designed for two essential tasks for which encryption is typically used:

• general encryption, used to protect information exchanged across a public network

• digital signatures, used for identity authentication.

Learn more on NIST PQC actions in this post:

Crypto Industry

The Bitcoin Ecosystem recognizes the threat posed by Quantum computing and is proactively working to address threats from Quantum computing to the current implementation of Elliptic Curve Digital Signature Algorithm (ECDSA) based signatures in Bitcoin. ECDSA creates the encryption keys used to protect Bitcoin wallets as well as playing a central role in securing crypto transactions and verifying the authenticity of transactions.

Bitcoin Improvement Proposal 360 (BIP-360) proposes adoption of Pay-to-Merkle-Root (P2MR) outputs that operate in similar ways to the current Pay-to-Taproot (P2TR) outputs but removes the “key path” spending option that exposes public keys on the blockchain, which Quantum computers could potentially crack. Test implementations of M2PR are now underway, and there is not yet formal timeline for adoption.

BIP-361, a companion Bitcoin wallet migration plan, suggests a multi-year rollout post-BIP-360 activation. It bans legacy address (using ECDSA) sends after 3 years, invalidating old ECDSA signatures after 5 years – essentially “freezing” a user’s bitcoin tokens. Analysts estimate 3-7 years total for full Quantum resistance if activation of BIP-360 begins soon, aligning with industry projections of meeting the 2029-2030 targets for completing preparations for Q-Day.

The Ethereum Foundation has prioritized post-Quantum cryptography to counter Quantum computing threats that could compromise elliptic curve signatures and other core mechanisms. The strategy involves dedicated teams, research funding, and a series of protocol upgrades via hard forks.

A post-Quantum security team was launched in early 2026 with dedicated funding to develop Quantum-resistant tools like hash-based signatures and to integrate NIST standards such as CRYSTALS-Kyber and Dilithium. The upgrade roadmap currently maps out four hard forks with Layer 1 upgrades targeted for 2029 and full execution layer migration taking additional years as follows:

• Fork I: Equips validators with activatable Quantum-safe public keys for emergency response.

• Fork J: Lowers gas costs for verifying post-Quantum signatures.

• Fork L: Compresses blockchain state into zero-knowledge proofs for efficiency.

• Fork M: Secures Layer 2 networks against Quantum attacks.

Learn more about Bitcoin and ECDSA encryption here:

Economic Advantage + Technological Leadership

As mentioned in the intro, Quantum computing is clearly identified as a dual-use technology – aka having with both civilian and military applications. This naturally pushes both nation states and commercial enterprises to implement “selected collaboration” where openness and sharing are balanced against control and protection of strategic advantage.

Export Controls and Protective Regulations - Balancing Collaboration with Competition

The U.S. has tried to restrict exports of key Quantum technologies such as precision lasers and ultra-low-temperature dilution refrigeration systems, cryogenic electronics, and phonics technologies.

Similarly, U.S. allies including Australia, Canada, France, Japan, Netherlands, and the UK have implemented expanded export controls to cover certain cutting-edge Quantum technologies with the objective to slow proliferation to nations of concern - primarily China, Russia, Iran, and North Korea.

Not surprisingly Chinese Quantum labs are working to source equipment domestically and the country is investing heavily. Also, not surprising, some analysts are posing that export controls are accelerating the development of a localized Quantum supply chain in China – aka the containment strategy being used by western countries may be producing exactly the outcome it was designed to prevent – not by blocking China’s progress, but by incentivizing China to build alternatives that the rest of the world can adopt.

At the same time, among friendly blocks of nations, we see increasing collaboration and an easing export control among these countries so they can share Quantum R&D more freely. This is helping to facilitate the creation of trusted supply chains where Quantum hardware, software, and knowledge can flow with fewer bureaucratic hurdles, ensuring the allies move faster collectively. The EU is working to harmonize rules to let researchers and companies in member states collaborate more easily.

Tightening Investment Screening – Walking Yet Another Fine Line to Walk

For countries, a funding challenge is finding the right balance between control/screening and under analysis. Over-regulate and you potentially stifle innovation. Under-regulate, and it could lead to a promising Quantum breakthrough becoming owned or accessible to a strategic rival – possibly creating a new risk.

Many countries view access to funding/capital as a strategic control vector to fend off adversaries versus supporting home-grown solutions. Allies are also coordinating, for example, the U.S.-EU Trade and Technology Council is discussing aligned approaches to screen investments and sharing information on concerning investors who may be trying to leverage loopholes or to sidestep restrictions.

The European Union has begun applying its Foreign Direct Investment (FDI) screening regulation that requires all member states to screen foreign investments in sensitive sectors like defense, semi-conductors, AI and Quantum technologies. FDI enables member States to assess, investigate, authorize, condition, prohibit, or unwind foreign investments when national security or public order could be affected.

What about the Wet-Ware (human beings)?

Currently, the demand for Quantum specialists worldwide outpaces supply, with estimates suggesting that only one out of every three Quantum job openings will be filled by 2025 in the U.S. alone. Combined with the concern that sensitive information and IP needs protection and controls on distribution, the human factors in the Quantum industry are complicated. The U.S. and allies are increasingly scrutinizing joint research projects and talent flows that might transfer critical knowledge to perceived adversaries.

Sharing certain software or technical data with foreign nationals may require an export license under U.S. law. This is creating a situation where public and private organizations working on Quantum technologies need to carefully assess research security and to carefully monitor compliance to prevent unintended leakage or the sharing of sensitive/restricted IP. Europe is working on similar initiatives to screen investments and partnerships in sensitive areas of Quantum technologies

In addition, the recruitment and retention of human talent for Quantum has become a tool of geopolitics. Immigration policies are no longer viewed in isolation from national security or technological primacy – it’s a front-line instrument in the quest for accelerating the development of Quantum technologies. Visa policies have become a primary battleground with countries friendly to innovation lowering barriers for skilled migrants, while others risk eroding their talent base through visa restrictions.

In short, what is needed is an intelligent mix of immigration policies that view human capital strategically and that balance open doors to welcome global expertise with selective restrictions to protect vital knowledge and IP.

Quantum Supply Chains

The status of supply chains for Quantum technologies can be characterized as nascent, thin, fragile, and having complex globally dispersed dependencies across materials, fabrication, cryogenics, lasers, control electronics/software, and specialized talent. In many areas commercial–scale capabilities are often unavailable, supplier bases are small, and scaling remains difficult across hardware, software and human talent.

An industry survey found that nearly 60% of respondents expected some Quantum–related supply chain disruption within three years, with raw materials and manufacturing equipment ranking as the leading causes. DownLoad the survey here (PDF):

The variety of Qubit modalities used in Quantum creates their own unique requirements, and it is no surprise that the various modalities have supply chains that have evolved into unique niches. The supply chain profile is not uniform across superconducting, trapped–ion, photonic, and semiconductor spin platforms. Each depends on different combinations of materials, fabrication steps, and subsystem suppliers. This makes supply chain resilience harder than conventional industrial planning, because there is no single stack that works across the field.

As nations and companies race to build Quantum computers, hardware components (specialized lasers, cryogenic systems, microwave electronics, etc.) become critical assets this can contribute to a potential risk vector. For example, if a country is procuring Quantum systems or components from other nation could there be backdoors or hidden vulnerabilities inserted into these systems that would allow an adversary to disrupt the operation of the Quantum computer or steal information from it?

This might sound a bit far-fetched, but it’s a valid concern in a world where tech supply chain integrity is already a big issue in Information/Data Processing and Communications/Networking equipment for example. Ensuring trusted suppliers and developing domestic Quantum hardware capabilities is on the radar of most governments to try to avoid developing dependencies on potentially compromised components. This in turn can further fragmentation and the creation of difficult to scale niches in supply chains.

Hardware supply chains are only part of the bottleneck. As mentioned above, access to technical expertise in hardware and software design/production is also a top challenge. That suggests the sector’s constraints are increasingly about execution capacity—people, process knowledge, and building scalable manufacturable workflows— and not about funding or building functioning prototypes in labs.

Implications for Military Power

Then there are the military dimensions for Quantum, which are numerous, and we will touch on a couple of them here. It is natural/logical for Quantum technology to be compared to the advent of nuclear weapons in terms of its potentially game-changing impacts on humanity. A powerful Quantum computer will be able break today’s data encryption codes/systems creating the very real possibility of crippling secure communications globally – essentially the modern equivalent of cracking the Enigma coding machine in World War II, but on a truly global scale.

Militaries foresee that Quantum computers will aid in weapons design, logistics, and AI. Quantum sensors could detect stealth aircraft or submarines by sensing minute anomalies in gravity or magnetism. Quantum networking leverages the unique properties of Quantum mechanics of superposition and entanglement to transmit Quantum information (qubits) between devices. For a Military, Quantum communications promise ultra-secure links with unbreakable encryption – critically important for command and control. Quantum networks hold the potential of creating a “Quantum Internet”.

Get the deep dive on Quantum Networking in this TechAptitude post:

It’s no surprise that Quantum technologies are now recognized as an arena of rivalry like that in the Cold War race to develop nuclear capabilities. In short, Quantum is viewed as a critical future military enabler, so much so that it appears alongside AI and Hypersonics in defense strategy documents of all major powers. The U.S.A. and China have explicitly framed Quantum as a domain of rivalry similar to the Cold War race over nuclear capabilities. This time, the “bomb” is a Quantum Qubit where the winner of the Quantum “race” could gain a decisive edge.

International Collaboration and Treaties

The dual use nature of Quantum technologies suggests that even if there are one or more “winners” in the Quantum race – no technical monopoly lasts indefinitely. Prime examples – the U.S. monopoly on the nuclear bomb lasted about 4 years. In the space race, Russia’s Sputnik ended any U.S. lead shockingly in 1957, and yet the U.S. became the first and only nation to land human beings on the moon. It is very possible that a lead in Quantum computing currently held by one nation or another will likewise be transient and relatively short lived.

Advances in physics as borne out by the Cold War are difficult/impossible to contain to a single nation state. Breakthroughs in physics tend to diffuse across geo-political boundaries, either through spying, IP leakage, independent discovery or simply through collaboration by researchers in multiple countries.

The international treaties and non-proliferation regimes that resulted from the Cold War will likely provide frameworks for the creation of new risk management regimes and agreements on the application of Quantum technologies to offensive military use cases and other high-risk areas like busting encryption algorithms to avoid dangerous escalation.

While nation-states are currently collaborating mainly through research alliances, standards-setting, regulatory guidance, and capacity-building, there is no comprehensive global regime or set of formal mechanisms in place to support multi-lateral conversations and deliberations about addressing the risks of Quantum computing. That said, there appears to be enough international coordination to show that governments recognize the threats are transnational and cannot be handled by any one country alone.

Currently these collaborations compete with national-security concerns and the “Quantum race” mindset, which can limit openness and slow progress. In practice, this means the international response is still fragmented: strong on research partnerships and early standards work, weaker on binding enforcement or globally coordinated migration deadlines.

Many countries appear to be using a practical and “selective collaboration” model where open cooperation in basic research, standards, and workforce development is supported, but tighter limits and controls on sensitive hardware, algorithms, funding, and defense-relevant applications are also being pursued. This selective collaboration approach tries to preserve innovation while reducing IP leakage and strategic exposure. A good example of this is the efforts around establishing Quantum-safe communication technologies where groups of allies coordinate standards and migration plans, while still guarding the most sensitive implementation details and supply-chain dependencies.

Wrapping UP!

While it is easy to suggest that “winning” at Quantum computing represents a ticket to dominance for one country over another. Clearly a “win the race” mentality is being used as justification inside countries for significant investment into the creation of well-coordinated Quantum strategies and robust research ecosystems so as to remain best-positioned to dominate the field. The impacts of Quantum technology will be hugely transformative across all aspects of the human existence – this no one doubts! The simple notion of “winning” is too simplistic and short sighted a viewpoint and it fails to capture the vast potential of future implications of Quantum Technologies – IMHO. 😊

Please, share your thoughts via the Comments section for this post or open a new SubStack chat thread … and please forward this post to your friends and colleagues. Until next time!

Let’s Get Started!

A Quick Overview of NASA Artemis

Formally established in 2017, NASA’s Artemis program is a Moon exploration program that aims to land humans on the Moon for the first time since Apollo 17 in December 1972. It also intends to establish a permanent base on the Moon in the 2030s. Core components in the program include the Space Launch System (SLS), Orion Spacecraft (the bit that holds human beings), and the Human Landing System (HLS). Many of the key hardware components used in Artemis were developed in previous NASA programs including the RS-25 engines and solid fuel boosters (Space Shuttle), Orion Spacecraft (Constellation Program).

Artemis leverages numerous primary contractors (Lockheed Martin, Northrop Grumman, SpaceX, Blue Origin, Aerojet Rocketdyne (L3Harris)) for core systems like the SLS rocket, Orion spacecraft, and lunar landers and uses secondary contractors (Amentum, Axiom Space, Bechtel, Intuitive Machines, Lunar Outpost, Maxar Space Systems, Teledyne Brown) to design, build supply specialized components like space suits, lunar vehicles and ground control systems.

The Artemis I mission sent an uncrewed Orion Spacecraft to orbit the Moon in 2022, and in April 1-10, 2026, Artemis II sent four astronauts on a lunar flyby around the moon. Artemis III is scheduled for 2027 with a focus on executing docking tests with an HLS lunar lander in Earth orbit. The first lunar landing of the Artemis program is planned for the Artemis IV mission, sometime in 2028.

SLS - Stages and Core Components

Development of SLS began in 2011 as a congressionally mandated replacement for the retiring Space Shuttle and the cancelled Ares I and Ares V launch vehicles of the Constellation program. SLS is a super heavy-lift expendable (aka it is used once and only once) space launch vehicle capable of lifting over 95 metric tons into earth orbit, with a Block 1 configuration standing 322 feet tall. The vehicle incorporates hardware from the Shuttle program, including its RS-25 engines and solid rocket boosters, integrated with a new SLS core stage.

Core Stage

Built by Boeing, SLS’s 212-foot-tall and 27.6 feet in diameter core stage is the primary gas tank to feed SLS’s four RS-25 engines. Core stage holds 733,000 gallons of liquid propellant – with separate liquid oxygen (LOX) and liquid hydrogen (LH2) tanks. SLS engines consume ~ 1,500 gallons of propellant per second during the eight minute liftoff sequence which the amount of time it takes SLS to reach orbit, at which point the Core Stage separates from the upper stage and NASA’s Orion spacecraft.

The stage is constructed from 10 major barrel sections, four dome sections, and nine rings. Each cylindrical barrel section consists of eight aluminum panels that vary in length and height. The aluminum panels are welded to form the five major physical sections: engine section, LH2 tank (5 barrels, 2 domes), intertank, LOX tank (1 barrel), and forward skirt, all constructed from welded aluminum-lithium panels. The core stage is covered with an orange spray-on foam to insulate the cryogenic propellants.

LOX – serves as the furl oxidizer in SLS, and is cryogenic oxygen liquefied at around -297°F (-183°C) and stored in the SLS core stage’s upper tank, which holds about 196,000 gallons (741,941 liters) weighing roughly 1.86 million pounds. LOX is non-toxic but reactive, making it a preferred, cost-effective oxidizer for large propulsion systems despite its cryogenic handling requirements.

LH2 – is the primary fuel in SLS core stage. LH2 provides exceptional efficiency and delivers superior exhaust velocity compared to other fuels due to hydrogen’s light mass and high energy release. For SLS it is cryogenically cooled to around -423°F (-253°C or 20 K). Cooling enables more propellant mass—up to 10% extra—to be stored in the core stage tank, and it makes LH2 low-density and low viscosity which aids smooth flow through lines. LH2 does require careful handling to prevent boil-off from ambient heat.

Core stage also includes the Main Propulsion System (MPS) with hydraulic gimbal actuators, pneumatics, helium pressurization, and a Cryogenic Auxiliary Power Units (CAPUs). Avionics in the forward skirt and intertank manage flight control, navigation, and communication during the active burn sequence.

Hydraulic Gimbal Actuators: RS-25 engines are a gimbaled thrust system aka the exhaust nozzle of the engine can be swiveled on two axes (pitch and yaw) enabling precise steering of the rocket. The Gimbal Actuators are Type III hydraulic servo-actuators (one per axis), that generate 3000 psig hydraulic pressure via pumps driven by gaseous hydrogen turbines

Cryogenic Auxiliary Power Unit (CAPU): The CAPU powers the hydraulic systems for thrust vector control enabling gimballing of the four RS-25 engines and control of engine propellant valves. Each CAPU (there are four CAPUs, one per RS-25 hydraulic system) includes a turbine wheel which is spun by pressurized gaseous helium during vehicle startup, then by hydrogen gas, a gearbox to reduce shaft speed, a gearbox heat exchanger, a propellant supply valve for on/off gas flow, and a 2-stage speed control valve.

The SLS uses a conical frustum-shaped interstage known as the Launch Vehicle Stage Adapter (LVSA) between the core stage and the narrower diameter upper stage. The LVSA consists of sixteen aluminum-lithium panels made of 2195 aluminum alloy and is built by Teledyne Brown Engineering.

RS-25 Engines

Per NASA, the four RS-25 engines attached to the underside of SLS’s core stage are “the most efficient engines ever built” and are expended after use. The first four Space Launch System flights use modernized and refurbished engines built for the Space Shuttle program. Subsequent flights will make use of a simplified RS-25E engine called the Production Restart.

RS-25E units which are currently undergoing testing and development and are faster to build and cost ~ 30% less than the originals. The RS-25E is rated at 522,000 pounds-force (lbf) or 2,321 kilonewtons (kN) thrust per engine – providing slightly more than over two million pounds of thrust total for the SLS. The first test firing of an RS-25E occurred in June 2025 and was declared successful.

Obviously, the RS-25 engines are highly sophisticated systems consisting of pumps, valves, fuel systems, and other components working in concert to produce thrust. Space constraints prevent us from going into great detail of all the sub-systems. Here is an overview of the core components.

Fuel System

HL2 Fuel and LOX oxidizer from the rocket’s core stage flow directly into the Main Propulsion System (MPS) lines, otherwise known as the “plumbing” feeding the engines. Once in the MPS lines, the fuel and oxidizer each branch out into separate paths to each engine where pre-valves allow the propellants to enter the engine. Once in the engine, the propellants flow through different paths containing several turbopumps, pre-stage coolers, heat exchangers, and pressurization before reaching the Main Combustion Chamber (MCC) injectors where the propellants are mixed and injected into the MCC where they are ignited.

Main Combustion Chamber

Fuel-rich hot gas from a hot-gas manifold enter the chamber at the injector with the mixture ignited by the “Augmented Spark Igniter”, an H2/O2 flame at the center of the injector head. The MCC is composed of a structural shell made of Inconel 718 which is lined with a copper-silver-zirconium alloy called NARloy-Z. The temperature in the combustion chamber reaches 3,300 °C (5,970 °F) during flight – higher than the boiling point of iron. NASA is researching the use of advanced structural ceramics, and ceramic-matrix composites for use in the MCC to create more efficient combustion and to reduce cooling requirements.

Engine Controller

Each engine has an integrated Main Engine Controller (MEC), which controls all the engine’s functions and monitors performance. Integrating the MEC directly into the engine greatly simplifies wiring between the engine and the launch vehicle, because all the engine sensors and actuators are connected directly to the controller. Each MEC is connected to the orbiter’s main computer and avionics system via its own dedicated interface unit making the system more reliable.

The controller contains two independent dual-CPU processors (four processors total). Automated failover will switch controllers if a problem is encountered. Within each system the two processors operate in lockstep, enabling each system to detect failures by comparing the signal levels on the buses within that system. Memory for the original controllers was plated-wire type, which functions similar to magnetic core memory and retains data even after power is turned off. Newer controllers will use conventional CMOS static RAM.

Main Valves

To control the engine’s thrust output, the Main Engine Controller operates five hydraulically actuated propellant valves on each engine, the oxidizer pre-burner oxidizer, fuel pre-burner oxidizer, main oxidizer, main fuel, and chamber coolant valves. A coolant control valve is mounted on the combustion chamber coolant bypass duct of each engine. The engine controller regulates the amount of gaseous hydrogen allowed to bypass the nozzle coolant loop, thus controlling its temperature.

Nozzle

The nozzle is a bell-shaped extension bolted to the main combustion chamber, referred to as a de Laval nozzle. Physically it is 121 in (3.1 m) long with a diameter of 10.3 inches (0.26 m) at its throat and 90.7 inches (2.30 m) at its exit. The inner surface of each nozzle is cooled by liquid hydrogen flowing through brazed stainless steel tube wall coolant passages.

Solid Fuel Rocket Boosters

Each of SLS’s two solid rocket boosters is the height of a 17-story building. Each generates 3.6 million pounds of thrust, providing 75 percent of the required power during the first two minutes of launch. The solid propellant combine oxidizers, fuels, binders, and additives to achieve desired performance characteristics like specific impulse and burn rate. Solid propellants come in two main types:

• “Composites” (the most common type) are mostly a mixture of granules of solid oxidizer, such as ammonium nitrate, ammonium dinitramide, ammonium perchlorate, or potassium nitrate in a polymer binding agent, with flakes or powders of energetic fuel compounds such as RDX, HMX, aluminum or beryllium. Plasticizers, stabilizers, and burn rate modifiers (e.g. iron oxide or copper oxide) can be added.

• “Double-Base” consisting of nitrocellulose and nitroglycerin (both providing fuel and oxidizer functions), often plasticized for stability, yielding simpler formulations but lower performance than composites.

NASA SLS uses a composite type of fuel centered on a PBAN binder system mixed with key oxidizer, fuel, and additives. It consists primarily of ammonium perchlorate (oxidizer, about 70% by weight), atomized aluminum powder (fuel, 16%), PBAN rubber (polybutadiene acrylonitrile binder and secondary fuel, 12-14%), iron oxide (catalyst, 0.2-0.4%), and an epoxy curing agent (about 2%).

• Ammonium perchlorate: Provides oxygen for combustion, making up the majority for efficient burning.

• Aluminum powder: the Fuel, delivers high energy density, chosen for its power and safety.

• PBAN binder: Holds the mixture together, also contributing as fuel.

• Iron oxide: Catalyzes the reaction to control burn rate.

• Epoxy curing agent: Hardens the propellant into a solid grain.

Interim Cryogenic Propulsion Stage (ICPS) – RL10 Engine

The ICPS, an engine built by Boeing and United Launch Alliance, provides in-space propulsion for the Orion Spacecraft to send on its trajectory towards the moon and back. The RL10B-2 engine will produce 24,750 pounds of thrust to power Orion towards the moon. Measuring 45 ft (13.7 m) tall and 16.7-foot (5.1 m) in diameter, it is a single-engine liquid hydrogen/liquid oxygen-based system that will fire up after the solid rocket boosters and core stage have served their purpose and are jettisoned.

Orion SpaceCraft

NASA’s Orion spacecraft is the crew vehicle for the Artemis program, designed to carry astronauts beyond low Earth orbit to the Moon and eventually support Mars missions. Orion includes a crew module and a service module. The Service Module is provided by the European Space Agency.

The crew module can accommodate up to four astronauts for about 21 days, featuring advanced life support, avionics, and an ablative heat shield capable of withstanding reentry velocities over 25,000 mph. Solar panels on the service module generate power, while its single RL-10 engine enables orbital maneuvers and deorbit burns.

The Service Module supports the Crew Module from launch through separation before re-entry. It delivers essential propulsion, power generation, thermal control, and life support consumables like oxygen, water, and nitrogen. It also includes Orion’s propulsion system, enabling orbital maneuvers, attitude control, and trajectory adjustments with a total of 33 maneuvering and thruster engines: one main Orbital Maneuvering System engine (OMS-E), eight auxiliary R4D-11 engines for primary thrust, and 24 reaction control system thrusters for fine maneuvering in all directions

Artemis II, launched April 1, 2026, marked the first crewed flight using Orion. The mission was a 10-day flyby of the moon with the goal of testing human-rated capabilities of Orion. Orion has the following core capabilities:

• Crew Safety: Autonomous abort systems, radiation sheltering

• Mission Flexibility: Supports lunar orbit, flybys, and future Moon gateway station operations

• Life support system for extended mission durations

• Automated docking system

• Glass cockpit interfaces aka windows

• A tower launch escape system

Artemis III will test Earth-orbit docking of Orion with a lunar lander like starship HLS, paving the way for 2028 landings on the Moon.

Fun Facts

• Each SLS solid fuel booster stands 177 feet tall, 12 feet in diameter, and weighs 1.6 million pounds

• Core Stage Empty Weight: Approximately 188,000 lbs. (85,275 kg)

• Core Stage Tank Capacities: Liquid Hydrogen (317,000 lbs) 537,000 gallons (2 million liters) Liquid Oxygen (1.86 million lbs) 196,000 gallons (741,941 liters)

SLS Configurations

NASA had originally planned to build multiple configurations of the SLS – referred to a Block 1, Block 1B, and Block 2. Block 1 and Block 1B share the same core stage, solid rocket boosters, and RS-25 engines but differ primarily in their upper stages and payload capabilities.

Block 1 Configuration: As detailed in this article, this initial variant uses a core stage with four RS-25 engines, twin five-segment solid rocket boosters (SRBs), and an Interim Cryogenic Propulsion Stage (ICPS) upper stage. Block 1 remains operational for the first three Artemis missions (I-III).

Block 1B Configuration: Planned upgrades with an Exploration Upper Stage (EUS) replacing ICPS, enabling larger payloads like habitats or landers, boosting capacity to 38-46 metric tons to translunar injection and 105 metric tons to low earth orbit.

Block 2 Configuration: Planned upgrades with advanced boosters (beyond five-segment SRBs), enhanced RS-25 engines, and EUS or equivalent, targeting 46+ metric tons to the Moon or 130 metric tons to low earth orbit.

Wrapping up

For sure, NASA SLS is an engineering marvel and we can expect to see it take human being to the moon in the next couple of years. STAY TUNED!

Please, share your thoughts via the Comments section for this post or open a new SubStack chat thread … and please forward this post to your friends and colleagues. Until next time!

Product Updates April 2026

• Apple Vision Pro: MR headset, available via Apple Stores/online; limited stock, Apple indicates Vision Pro 2 will come out in 2026.

• Google Glass: no longer sold or supported.

• HTC Vive Focus Vision: MR headset, consumer product available.

• Magic Leap 2: AR headset, Developer Pro version available, intended for enterprise use cases.

• Microsoft HoloLens: is discontinued and out of stock.

• RayNeo Air 3s: AR glasses, consumer product available.

• Samsung Galaxy XR: MR headset, consumer product available.

• Snap Spectacles: AR glasses, developer-only, available via monthly lease.

• Sony PSVR2: Tethered VR headset for gaming, consumer product available.

• Viture Pro XR: AR glasses, consumer product available.

• Xreal One Pro: AR glasses, consumer product available.

Notice that many of the products listed above, at least those that are still available, are AR products not VR capable headsets.

The real BIG story is META and the change in product strategy re the MetaVerse!

In March of 2026, Meta announced it would remove Horizon Worlds from the Quest VR store by late March and fully end VR access by June 15, transitioning to it’s mobile-only app. Previously in Jan 2026, Meta laid off more than over 1,500 staff in its Reality Labs unit. Meta then quickly reversed course indicating that existing VR worlds would remain accessible, but that no new VR content or development is being planned.

Recall that Meta Reality Labs is the division focused on advancing virtual reality (VR), augmented reality (AR), and mixed reality (MR) technologies. It originated from the 2014 acquisition of Oculus VR and has since expanded into a major R&D unit driving Meta’s MetaVerse vision and key products including Quest series of VR/MR headsets, and Ray-Ban Meta smart glasses.

Reality Labs has racked up over $80B in losses since 2021, and experienced underwhelming user adoption which has created investor pressure to de prioritize the MetaVrse. Meta is now closing in-house studios and treating Horizon Worlds as a low-priority relic rather than a core product bet.

Meta’s current strategy for Reality Labs emphasizes a pivot from heavy MetaVerse investments toward AI integration, focusing on wearables like Ray-Ban Meta smart glasses, and sustainable VR development. Meta continues to claim they are maintaining VR headset development. Meta is now positioning wearables and AI as the entry points for “immersive computing”.

Please, share your thoughts via the Comments section for this post or open a new SubStack chat thread … and please forward this post to your friends and colleagues.

Continue below to read the previous iterations of this post.

++++++++++++++++++++++

Updates Sept 2025

++++++++++++++++++++++

I wrote the original version of the post in 2018 when assorted firms were determined to build and bring to market immersive Virtual Reality headsets. I firmly believed then that the pursuit of Virtual Reality products was destined to be a #Fail, and IMO, that has been borne out. That said, VR headsets have been largely pushed aside by new and “exciting” AR glasses offerings.

Here are a few recent developments in this space to consider:

Apple Vision Pro: a mixed-reality spatial computer first announced in June 2023 and released in early 2024 has been commercial disappointment with total sales of less than 500K units. Apple has teased future products including Vision Air and Apple Smart Glasses that may see the light of day in 2027 or 2028.

Meta Ray-Ban Display: Sept 2020, Meta teased the release of a Ray-Ban glasses product with the goal of creating ”full augmented reality”. This was part of the larger push by Meta to create the “MetaVerse”.

• Sept 2021, Meta released the Ray-Ban glasses line, called “Ray-Ban Stories”.

• Oct 2023, Gen 2 release of Ray-Ban glasses Skyler, Wayfarer and Headliner models offering an improved camera and microphone, water resistance, and a voice interface. June 2025 announced the “Oakley Meta” glasses. Sept 2025, Meta announced “Meta Ray-Ban Display” – AI glasses with an integrated full-color, high-resolution display built into the right lens. The high profile product announcement was marred by live in person demos that failed miserably.

• Meta’s Ray-Ban and Oakley products have been offered at a variety of price points and in spite of doubling and tripling down on these products and spending estimated tens of billions of dollars Meta is yet to gain a significant return. Estimates are that Meta has sold 2 million units since launching their second generation in 2023, with Meta’s optimistic projections of 5 million units being sold in 2025.

Meta Quest/Oculus VR: Originally developed by Oculus VR these products include several generations of headsets providing immersive VR experiences ranging from PC-tethered to standalone models. Now marketed as “Meta Quest” and targeted to Metaverse experiences with new versions originally slated for 2026, now reportedly delayed until 2027.

Microsoft: Began development of AR glasses with HoloLens and IVAS headsets in 2018 for the US Army. The project goal was to equip over 120,000 soldiers with headsets that could overlay real-time battlefield information, such as thermal and night vision, threat alerts, 3D terrain maps, and live streaming from weapon scopes and sensors. In 2024 Microsoft partnered with Anduril Industries who took over development, manufacturing, and ongoing hardware and software improvements of the IVAS system.

Snap Spectacles: Were first released in 2016 and now in generation 5. Spectacles are offered primarily as a developer and creator tool, not a mass-market consumer device.

Google Glass: Originally released in 2013, Google stopped consumer sales in 2015 and shifted focus to “Glass Enterprise” products targeted at professional markets such as healthcare, manufacturing, and warehouse logistics and products were released in 2019 and discontinued in 2023. Google is now working on Android XR, an augmented reality operating system for glasses announced in 2025.

Here is THE key challenge/problem with all this stuff …. The simple lack of a COMPELLING use case users care about! At least care about enough to ditch their other smart phone companion devices - aka watches, ear buds ….etc. Vendors keep dumping more and more tech (can you say AI anyone) with the hope that premium tech is enough to motivate buyers to pony up hundreds and hundreds of dollars for “cool”.

Video recording and integrations with health-tracking software and letting the user see text, images, and video overlaid onto their field of vision with glasses ain’t going to cut it as a sustainable value add to existing phones devices. Nor will wearing a headset as a Disneyland tour guide make the grade. Niche use cases YES – but these are NOT the transformational use case that everyone MUST have.

+++++++++++++++++++++++

Original Post Circa 2018

+++++++++++++++++++++++

Whaaatttt. You are kidding right? VR a failure? Yes indeed!

Virtual Reality (VR) is one of today’s brightest, shiniest objects. It is pitched and hyped as destined to alter human “reality” forever – ok – pun intended.

I submit - VR will NOT be the transformational success it is pitched to be. VR WILL be adopted in selected niche areas – yes, it is cool, but it will forever be “niche-y”.

To clarify – in stating VR will be a #Fail – I am suggesting VR will not achieve the level of mass adoption we have seen with things like the automobile and the “smart” phone where millions and millions of units are sold annually. Nor will VR have a societal impact like modern wide body commercial aircraft that altered the human experience on a global scale.

Here are some things to consider (in no particular order) suggesting VR may not be the ultimate slam dunk its purported to be. VR will be useful in some niches.

This VR stuff is not reality. Reality is multi-dimensional analog – and always will be.

VR forces the user to dedicate their attention onto what is being projected in front of their eyeballs. This is a big issue – IMO.

Humans are context switching animals. We like to switch focus and move our attention rapidly and in unique ways. Humans are very adept moving our concentration on different aspects of our environment on-demand. VR requires the user go all in - all the time – with a singular focus and this can be extremely taxing. This singular concentration can cause physical side effects too - note some VR users experience nausea. Other possible side effects range from sore eyes to seizures.

Any attempt to pay attention to the analog world outside the VR head display are next to impossible.

Forget taking a drink of a beverage, glancing at that incoming call on your mobile phone (oh wait – your phone is sitting in front of your eyes in its VR head band), or moving your attention to your crying baby. Switching context requires you break out of your VR session.

To be truly transformative, VR needs a SIMPLE neural interface much like the Na’vi Neural Connection in the movie Avatar. A simple connection mechanism where you plug in and are completely immersed into a new world with all of our natural senses (vision, taste, smell, touch, emotions) seamlessly supported.

VR needs to be completely immersive – today it’s a projected digital display operating in the analog world. Projecting onto your eyes won’t cut it. VR needs to connect into your central nervous system. This is not happening anytime soon.

Requires a banded projection device worn over the eyes - wireless or wired. Some require you to awkwardly insert a smart phone into a holder. Even the $199.00 Oculus Go or Magic Leap One devices will not change the need to place a projection device over the eyes.

Lack of standardization -proprietary and competing architectures. VR cannot scale to millions of users without standardized systems and standardized methods for creating and presenting content – technical challenges will continue to exist for many years.

It’s been coming soon for over 30 years now! I first learned about VR in 1980’s – read about Jaron Lanier’s foundational work on VR. Wikipedia post on Jaron Lanier is here. Better yet – learn about and read his books!

There is no killer use case or application. VR gaming is all fine and dandy – but that is just a niche. VR doesn’t address a mass market need or use case – it is not solving a universal problem – its primarily focused on entertainment uses, so by definition is a niche application.

The label “Virtual Reality” is a bit of a misnomer. How about a new term like immersive simulation. After all VR as currently constructed will always be a digital simulation that mimics the real analog world.

Ultimately #Success for VR will take massive consumer adoption driven by a transformative use case. Much of the above analysis looks at consumer implementations versus the application of VR in industrial or health care use cases. Surely there are some VR successes out there?

There are reports that some targeted industrial and medical use cases have been successful. Presumably many people have benefited from complex surgical procedures that were originally designed with VR simulations. VR will ultimately develop niches of success.

Yes, VR is cool stuff. Lots of new tech is cool stuff. Lots of cool tech stuff #fails! VR as currently implemented will always be a digital overlay into our analog reality. It requires singular concentration by the user which limits the time many users can devote to the experience.

Vendors and pundits will continue to claim VR is successful – they want to sell you VR gear! After 30 years of effort and investment we are still stuck in the early hype phase for VR. It is and will be a niche solution for a few select areas like gaming, product demos and medical procedure training. Because VR lacks a universal killer use case that will cause my grandmother to use VR – it has to be a #Fail.

End.

++++++++++++++

Let’s get started!

Starting Point - Recall what is a Qubit?

As detailed in this Quantum Tech Overview Article, the basic unit of information in quantum computing is the qubit, or quantum bit. Qubits are essentially the quantum equivalent of the traditional bit used by classical computers to encode information. What makes qubits special is they can behave like a bit and store either a zero or a one, or a qubit can also be a weighted combination of zero and one at the same time (called superposition) which makes the scope of new computational possibilities massive.

IBM – NightHawk

IBM has been a long time player in the Quantum space – having been researching quantum computing since the 1970s. The company regularly publishes a comprehensive roadmap with milestones it is working to achieve. IBM introduced Quantum System One in 2019, the world’s first integrated quantum computing system designed for commercial use. This 20-qubit system marked a significant step toward making quantum computers viable for business applications.

Most recently, in late 2025 IBM announced its most advanced processor yet – NightHawk and indicated it should be available to users via the cloud and in on site systems by the end of 2026. Nighthawk delivers 120 superconducting transmon qubits in a square lattice connected by 218 tunable couplers, a 20% jump in connectivity over its predecessor, Heron. This layout boosts entanglement and circuit depth without spiking errors, supporting up to 5,000 two-qubit gates.

IBM fabricates transmon qubits on silicon substrates using superconducting materials like niobium or tantalum, with lithographic techniques compatible with existing semiconductor processes, including recent 300mm wafer fabs.

Transmon Qubits – these are tiny superconducting electrical circuits that behave like an artificial atom and can store a quantum bit of information. They are formed with two superconductors separated by a very thin insulator, which acts like a special, nonlinear inductor – called a Josephson junction.

The chip is cooled to extremely low temperatures (around a few thousandths of a degree above absolute zero) so the metal becomes superconducting and quantum effects dominate. Transmon qubits are a successful and widely adopted qubit design in superconducting quantum computing

Tunable couplers are specialized devices in superconducting quantum processors, that control interactions between qubits on demand, supporting qubits to connect precisely when needed for computations, then disconnect to prevent unwanted interference. Nighthawk supports cloud access to NightHawk with more than 100-qubit processors and integrates with IBM’s Qiskit software for algorithm development.

Oiskit Software Stack

Qiskit is an open-source Python-based software stack for quantum computing supporting users to create, optimize, and execute quantum programs on real hardware or simulators. Oiskit implements a modular, extensible architecture with core components like circuit builders, transpilers, primitives, and optimizers, allowing scalable workflows from high-level abstractions to low-level gates and it supports backend-agnostic execution across providers like IBM, IONQ, and AWS.

Qiskit provides tools for building quantum circuits with gates and libraries, simulating executions, and running on cloud hardware via Qiskit Runtime. It includes AI-powered code assistance, a Functions Catalog for pre-built algorithms in optimization and chemistry, and Serverless for hybrid quantum-classical supercomputing. It also includes visualization, benchmarking via Benchpress, and ecosystem extensions.

Core Technologies

• Qubit type: 120 superconducting transmon qubits fabricated on silicon substrates

• Operating conditions: requires cryogenic cooling to near absolute zero temperatures in dilution refrigerators

• Control and readout: leverages FPGA-based electronics for pulse generation, with real-time error decoding under 480 nanoseconds

• Coherence times: achieved over 400 microseconds

IBM has plans to scale Nighthawk to deliver up to 7,500 gates by the end of 2026, followed by up to 10,000 two-qubit gates in 2027, and by 2028 potentially up to 15,000 gates.

IBM’s Hybrid Model for Quantum+Classic Computing: Quantum-Centric Supercomputing (QCSC)

Leaving no stone unturned, IBM clearly understands that Quantum Computing will exist in a world now dominated by classic 1’s and 0’s based computing. In March 2026, IBM published a reference architecture or blueprint (no pun intended 😊 ) to illustrate how it may be possible to bring quantum and classic computing together to run workloads in what IBM calls Quantum-Centric Supercomputing.

The three layer architecture maps a base hardware infrastructure. The base is the quantum system, with one or more interconnected Quantum Processing Units (QPU), a classic compute runtime using specialized classical FPGAs, ASICs, and CPUs whose job is to enable QPU operations from error correction coding to qubit calibration to active qubit reset. The second tier includes programmable CPU and GPU systems that are co-located with the quantum system and connected via a low-latency, near-time interconnects. The upper layer is the orchestration layer that includes the Quantum Resource Management Interface (QRMI), an open source library that abstracts away hardware-specific details and delivers APIs for quantum resource acquisition, task running, and systems monitoring.

IONQ TEMPO - Trapped Ions

IONQ employs a trapped-ion approach, which is claimed to represent the most accurate quantum computing technology currently available, and it is implemented at room temperature. IONQ deploys barium ions as qubits - ionized atoms held in linear chains within custom ion traps. These traps use radiofrequency (RF) electrodes to suspend ions in an ultra-high vacuum for isolation from environmental noise.

Trapped-ion systems have become one of the established platforms for advancing quantum technology. These systems use electric fields to trap and move ions in a quantum processor, as well as lasers to manipulate their atomic and motional quantum states. This architecture supports the deployment of long chains of interconnected qubits that remain in a state of quantum coherence for long periods of time.

Qubits are stored in stable electronic states of each ion, and quantum information can be transferred through the collective quantized motion of the ions in a shared trap. Lasers are applied to induce coupling between the qubit states (for single qubit operations) or coupling between the internal qubit states and the external motional states for entanglement between qubits.

Like classical CPUs, the size of IONQ’s individual Quantum Processing Unit (QPU) is limited by the high-fidelity entangling gates that need to be deployed over the length of the chain which creates a practical size limit for IONQ’s QPUs – aka at some point connecting the qubits becomes too complex. To address the limitation IONQ is working to build multicore QPUs, that place multiple processing cores on one chip similar to how multicore CPUs used in classic computing.

Multiple ion chains are manipulated to dynamically form quantum computing cores – essentially the computing cores can be reconfigured to ensure the system utilizes the full computational capabilities of all the qubits in the system, thereby increasing the computational power exponentially.

Core Technologies

• Qubit type: supports 100 qubit trapped-ion quantum computing in a single ion chain with all-to-all connectivity, and it leverages individual barium ions

• Operating conditions: operates at room temperature

• Control and readout: control systems leverage precisely calibrated lasers: individual beams for single-qubit rotations and global beams for entangling gates via ion chain vibrations

• Coherence times: IONQ claims industry-leading coherence: T1 of 10-100 seconds and T2 (phase coherence) around 1 second

IONQ’s approach is differentiated by its long coherence times and all-to-all qubit connectivity. We can expect IONQ to be a leader in the Quantum Computing field going forward.

D-Wave Advantage2 – Quantum Annealing

D-Wave utilizes Quantum Annealing running adiabatic quantum computing algorithms - a system optimized for solving complex optimization and sampling problems by finding low-energy states in defined energy landscapes.

The system uses superconducting fluxonium qubits in loops, where qubits start in superposition and evolve through a programmable Hamiltonian that gradually introduces problem-specific biases and couplings. Couplers between qubits enable entanglement, allowing correlated states (same or opposite) to influence outcomes. D-Wave’s Advantage2 system deploys more than 4,400 qubits in a Zephyr topology with 20-way connectivity. Qubits are oriented vertically or horizontally and use shifted connections via three coupler types with each qubit connecting to 20 others—sixteen via internal couplers plus two external and two odd couplers to aligned qubits.

Quantum Annealing is essentially a computational process used to find optimal solutions to complex problems. It is a method of computation inspired by the principles of quantum physics and operates on the concept of “annealing,” a heating and cooling process traditionally applied to metals. Quantum Annealing uses quantum effects instead of thermal energy.

It explores multiple possible solutions simultaneously through quantum effects, potentially finding high-quality answers to certain optimization problems more efficiently than traditional computing methods. It has the potential to handle problems of scale and complexity beyond traditional computing capabilities.

The Annealing Process

At the start of the annealing process, the qubit is in a superposition state, which can be represented as a single valley with a single minimum energy. As the Quantum Annealing process runs, an energy barrier is raised which separates the single minimum energy into two valleys – also called a double-well potential.

As the system evolves, the potential evolves into a double-well shape, meaning that at the end of the anneal cycle, the qubit can end up in one of the two valley states. The problem encoding controls the outcome of this process: the magnetic field bias will tilt the potential, and the couplers allow interaction between qubits in a way that leads to correlation of their measurement outcomes. By the end of the anneal cycle, qubits are in a state that ideally represents the minimum energy state of the encoded problem.

In Quantum Annealing, tunneling enables qubits to escape local energy minima—shallow traps or valleys—and reach the global minimum by “tunneling” through barriers between states. Local energy minima in Quantum Annealing refers to suboptimal states in the energy landscape where the system’s energy is lower than nearby configurations but higher than the global minimum.

Quantum Annealing leverages quantum tunneling to navigate between local minima toward lower-energy states. Quantum tunneling occurs when a particle passes through a potential energy barrier even if it lacks the classical energy to climb over it, due to its wave-like nature in quantum mechanics. Unlike a classical ball that bounces off a hill, the particle’s probability wave extends through the barrier, decaying exponentially but retaining a non-zero chance of emerging on the other side.

Adiabatic Quantum Computing

Adiabatic quantum computing algorithms rely on the adiabatic theorem of quantum mechanics to evolve a system’s ground state toward a solution. The adiabatic theorem states that a quantum system remains in its ground state if the Hamiltonian H(t) changes slowly enough compared to the energy gap between the ground and first excited states.

In practice, the system starts with a simple initial Hamiltonian shown as H_B, known as an easy ground state, and evolves to a problem Hamiltonian H_P whose ground state encodes the solution. Implementation steps are:

• Encode the problem into H_P, where the ground state minimizes an energy function matching the objective.

• Initialize in H_B’s known ground state.

• Anneal: linearly interpolate H(s)=(1-s)H_B+sH_P for s from 0 to 1 over time T and measure the final state for the answer.

Hamiltonian

The Hamiltonian of a system represents the total energy of the system; that is, the sum of the kinetic and potential energies of all particles associated with the system. The Hamiltonian takes different forms and can be simplified in some cases by taking into account the concrete characteristics of the system under analysis, such as single or several particles in the system, interaction between particles, kind of potential energy, time varying potential or time independent one. A time-varying Hamiltonian gradually shifts from an initial “driver” Hamiltonian that promotes mixing of states to a final “problem” Hamiltonian encoding the specific optimization task, like minimizing energy in a landscape of valleys and peaks.

Core Technologies

• Qubit type: ~ 4400 superconducting fluxonium qubits in loops

• Operating conditions: requires cryogenic cooling and operates at a temperature below 20 mK

• Control and readout: Control relies on multiplexed digital-to-analog converters (DACs). QPUs use several controls that are manipulated by individual on-QPU DACs.

• Coherence times: relaxation times in excess of 100 microseconds

Quantum Annealing faces several real world challenges including:

• Integration of quantum solutions with existing business systems and workflows - requires specialized expertise

• Unclear ROI where currently most business applications are experimental, making it difficult to predict when ROI will occur

• Problem formulation -translating actual business problems into formats suitable for quantum processing requires both quantum knowledge and deep domain expertise

While Quantum Annealing systems are still maturing, advancements in both theory and practical quantum infrastructure are bringing practical applications within reach such that Quantum Annealing could play a major role in revolutionizing many industries.

Wrapping Up

Quantum computing is a long-term endeavor that requires sustained funding and patience. Investors and governments are increasingly expecting measurable progress and clear timelines for return on investment. This creates tension between scientific uncertainty and business expectations. Over promising could risk disillusionment, while under promising may slow down investment support.

The brief overview of the three approaches outlined here shows this is not just about building quantum chips. Each company is betting on fundamentally different methodologies and technologies to solve the hardest problems in the quantum field. While great progress is being made, which of them will become an industry leading technology in practical hardware has yet to be determined.

Please, share your thoughts via the Comments section for this post or open a new SubStack chat thread … and please forward this post to your friends and colleagues. Until next time!

These are not iterative upgrades or speculative papers. Willow, Majorana 1, and Ocelot are real, physical chips. Each represents a distinct architectural vision, and a very different bet on the future of quantum computation. We plan to explore additional hardware designs in a future post – coming soon!

Let’s get started!

Starting Point - Recall what is a Qubit?

As detailed in this Quantum Tech Overview Article, the basic unit of information in quantum computing is the qubit, or quantum bit. Qubits are essentially the quantum equivalent of the traditional bit used by classical computers to encode information. What makes qubits special is they can behave like a bit and store either a zero or a one, or a qubit can also be a weighted combination of zero and one at the same time (called superposition) which makes the scope of new computational possibilities massive.

Microsoft Majorana 1 Processor – Topological Qubits

Instead of focusing on qubit volume, Microsoft’s Majorana 1 emphasizes qubit stability. It uses a novel material called a topoconductor, which enables the detection and control of Majorana zero modes. These quasiparticles could allow for very error-resistant qubits, a long-sought holy grail in the field.

Majorana zero modes use pairs of electrons and the topology of the electronic band structure to enable zero-energy quasiparticle states (exotic quantum states) that are localized at the system’s boundaries or defects. When you exchange (braid) these quantum states, the quantum state changes in a way that depends on the path, not just the final positions and this is key for topological quantum computing.

The topoconductor, or topological superconductor, is a special category of material that can create an entirely new state of matter – not a solid, liquid or gas but a topological state. This is harnessed to produce a more stable qubit that is fast, small and can be digitally controlled, without the tradeoffs required by current alternatives.

This breakthrough required developing an entirely new materials stack made of indium arsenide and aluminum, much of which Microsoft designed and fabricated atom by atom. The goal was to coax new quantum particles called Majoranas into existence and take advantage of their unique properties. These exotic particles possess special topological properties that make them nearly immune to environmental disturbances. Once thought to be purely theoretical, Microsoft claims it has now provided experimental confirmation of the existence of Majorana zero modes.

Topological quantum computing uses exotic “quasi-particles” and “knot-like” paths to store quantum information in a way that is naturally protected from many kinds of noise – making them very stable – in theory. Topological qubits store information in global, “shape-like” properties of a system, a bit like tying a knot in a rope. As long as you do not “cut” the knots, your computation is safe from many small errors. The knot “shape” is a braiding pattern of special excitations called anyons, often realized as Majorana zero modes.

Majorana quasiparticles arise in special nanostructures made of superconductors and semiconductors. Microsoft’s topological qubit architecture has aluminum nanowires joined together to form an H. Each H has four controllable Majoranas and makes one qubit. These Hs can be connected, to, and laid out across the chip like so many tiles. The resulting qubits are particularly resistant to decoherence with the storage of quantum information especially stable.

Majorana 1 Core Technologies

• Qubit type: Eight Majorana qubits in “topoconductor” heterostructures about 1/100th of a millimeter, or ~10 µm, in size. Chip is designed scale to millions of qubits.

• Operating conditions: Requires cryogenic cooling to near absolute zero temperatures.

• Control and readout: Custom CMOS electronics running at cryogenic temperature route and time‑multiplex many gate and measurement lines. Readout is executed with on‑chip resonators and coupling circuits integrated close to the qubit structures and leverages low‑noise and room temperature amplification and digitization.

• Coherence times – qubit parity lifetimes on the order of ~ 10-12 ms have been publicly reported.

Majorana 1 claims challenged

Industry researchers have questioned some of Microsoft’s claims suggesting that the tests used by Microsoft for detecting Majoranas are flawed and indicating the tests could be fooled by false positives, suggesting that Majoranas remain theoretical.

In summary, Majorana qubits are quantum bits encoded using Majorana zero modes, powered by exotic quasiparticles. A Majorana qubit uses pairs of Majorana zero modes to encode a single quantum bit of information in a delocalized, topologically protected way. This approach could enable qubits that remain stable much longer than conventional qubits and that can be manipulated via braiding operations for error-resistant quantum computing.

To date, no commercial or large-scale quantum computer uses Majorana qubits – all efforts are in the research stage. What started as theories in 2001 and 2010 have been turned into first evidence by 2012, and now into experiments getting closer to real qubit implementations.

Alphabet Willow – Transmon Qubits

Near‑term, Alphabet/Google is positioning Willow as a research platform to test scalable error correction and complex quantum dynamics. Longer‑term Willow is part of Google’s quantum roadmap to build large‑scale, fault‑tolerant quantum computers that could tackle chemistry, materials, optimization, and possibly AI‑related linear algebra use cases.

Chip Layout: Willow arranges its physical qubits on a 2D grid tailored for a “surface code error‑correcting” scheme, where data and “check” qubits are interleaved. Numerous physical qubits are combined into a single logical qubit, allowing errors to not only be detected but also efficiently corrected.

Real‑time decoding: Willow integrates a fast decoder that identifies likely error patterns and corrects them in real time for lower‑distance codes supporting continuous error‑detection cycles. Using advanced control software, Willow implements automated calibrations, machine learning for fine-tuning pulses, and a reinforcement learning agent to optimize error correction performance.

Each qubit is controlled with carefully timed microwave pulses that manipulate its quantum states, while measurement devices capture outcomes without collapsing the entire system. The lattice design reduces crosstalk between qubits and allows for more scalable error-correcting codes. This way, logical qubits can be preserved even when physical qubits fail.

Willow Core Technologies

• Qubit type: 105 Superconducting transmon qubits, tiny circuits that conduct electricity without resistance at ultra-cold temperatures.

• Operating conditions: The chip runs at temperatures near absolute zero (millikelvin range) to keep the qubits coherent and reduce noise.

• Control and readout: Each qubit is driven by precisely shaped microwave pulses to implement 1‑ and 2‑qubit gates, and dedicated measurement circuitry reads out qubit states without disturbing the whole device.

• Coherence times - Google reports qubit T1 coherence times “up to 100 µs” with public sources giving a spec of ~ 68 μs.

Transmon Qubits – these are tiny superconducting electrical circuits that behave like an artificial atom and can store a quantum bit of information. They are formed with two superconductors separated by a very thin insulator, which acts like a special, nonlinear inductor – called a Josephson junction.

The chip is cooled to extremely low temperatures (around a few thousandths of a degree above absolute zero) so the metal becomes superconducting and quantum effects dominate. Transmon qubits are a successful and widely adopted qubit design in superconducting quantum computing

Error Correction Overhead - One of the biggest hurdles for Willow—and for quantum computing in general—is the overhead required for error correction. Quantum states are fragile, easily disrupted by noise, heat, or imperfect control signals. To keep logical qubits stable, many physical qubits are needed to support just one error-corrected qubit. In Willow’s case, the lattice design has promising surface-code techniques, but it still needs a large qubit surplus for every fault-tolerant operation and scaling to millions of physical qubits is key before reaching useful, large-scale quantum systems.

‍Scaling Infrastructure - Even with Willow’s lattice design, building large quantum computers comes with major infrastructure challenges. Superconducting qubits require dilution refrigerators that cool to a fraction of a degree above absolute zero. As the number of qubits grows, so does the demand for more complex wiring, microwave control lines, and shielding to prevent interference. Expanding from a few hundred qubits to tens of thousands or even millions will require facilities that resemble data centers with extreme cryogenic systems. Power consumption, heat management, and physical footprint all become limiting factors.

Amazon Ocelot – Cat Qubits

AWS’s approach prioritizes reducing error rates first, and AWS has designed Ocelot specifically for error correction, using an approach that integrates error-resistant qubits directly into the hardware. Ocelot implements a clever combination of two types of qubits: “Cat Qubits” and “Transmon Qubits”. Cat qubits are a special form of superconducting qubit that intrinsically suppresses certain errors where a single quantum state is distributed over multiple photons. They are named after Schrödinger’s famous thought experiment involving a cat that is simultaneously alive and dead.

Cat Qubits - are quantum bits encoded in the states of bosonic oscillators – aka modes of a microwave cavity, which correspond to two opposite-phase oscillation states of the field. Instead of relying on a single two-level quantum element, a cat qubit stores information in two coherent states of a harmonic oscillator and their quantum superposition. This is analogous to two “classical” states (like a pendulum swinging to the right vs. to the left) that a harmonic oscillator can have.

Ocelot is a small-scale prototype chip designed to test the cat qubit approach. It is built with two integrated silicon microchips, (1 sq centimeter in size), bonded together in a stack. The chips contain superconducting circuits made from a thin film of tantalum, a material AWS scientists processed to improve its performance.

The key advantage is that these qubits exhibit a strong noise bias that reduces bit-flip errors. AWS reports bit-flip error times approaching one second – or over 1,000× longer than a normal superconducting qubit’s lifetime. These improvements in bit-flip error rates come with a price of increasing phase-flip error rates. To counter phase-flip errors, the design uses transmon qubits to act as error detectors, capturing phase-flip errors that can then be corrected. The biggest disadvantage of this method is that the transmon qubits themselves are prone to errors.

Ocelot Core Technologies

• Qubit type: Superconducting quantum chip with 5 Cat Qubit (data Qubits), 5 Buffer Circuits, 4 Transmon Qubits for error correction optimized for integration into Amazon’s cloud hardware stack.

• Operating conditions: The chip runs at temperatures near absolute zero (millikelvin range) to keep the qubits coherent and reduce noise.

• Control and readout: Uses special microwave circuitry (oscillator modes - on‑chip resonators) plus built‑in error correction

• Coherence times: AWS reports physical bit‑flip lifetimes approaching 1 second, with the tradeoff of phase‑flip times reported to be in the order of tens of microseconds, around 20 µs.

Cat Qubits represent a promising leap forward, but several challenges remain on the path to large-scale, fault-tolerant quantum computing with this approach. One major challenge is scaling up the correction of phase-flip errors. Another challenge is hardware complexity and stability. Cat qubits demand high-Q resonators and nonlinear couplers (for stabilization and for operations). The hardware is more complex than a simple transmon circuit, and with many elements (resonators, couplers, ancillas) per logical qubit, the system could be more susceptible to crosstalk, leakage, or fabrication variability.

Wrapping Up

Quantum computing is a long-term endeavor that requires sustained funding and patience. Investors and governments are increasingly expecting measurable progress and clear timelines for return on investment. This creates tension between scientific uncertainty and business expectations. Overpromising could risk disillusionment, while under promising may slow down investment support.

The brief overview of the three approaches outlined here shows this is not just about building quantum chips. Each company is betting on fundamentally different methodology and technologies to solve the hardest problems in the quantum field. While great progress is being made, which of them will become an industry leading technology in practical hardware has yet to be determined. According to experts, a quantum processor must have at least 1,000 qubits to be suitable for practical applications. We have a way to go to get to this milestone.

Please, share your thoughts via the Comments section for this post or open a new SubStack chat thread … and please forward this post to your friends and colleagues. Until next time!

In the post we will explore the core foundations of PWSA:

• Transport Layer: Provides secure, resilient, low-latency communication via an optical mesh network.

• Tracking Layer: Provides detection and tracking of advanced missile threats, including hypersonic vehicles.

DoD’s Shift to Optical Interoperability and Large Constellations of Small Satellites

PWSA will be deployed with optical/laser communications as the primary network/data transport layer by establishing interoperability standards for Optical Communication Terminals (OCT), enabling a resilient satellite based “mesh network” in Low Earth Orbit that can scale across multiple vendors and platforms. Net-net, laser links are the critical infrastructure in PWSA.

Unlike traditional radio frequency (RF) communications, which are susceptible to jamming and bandwidth limitations, Optical Inter-Satellite Links (OISL) provide a high-bandwidth, low-probability-of-intercept backbone.

PWSA Satellite Vehicles (SVs)

Further, to improve resilience and to meet the threats posed by strategic competitors, PWSA will deploy hundreds of small, highly networked and capable satellites aka “Space Vehicles” (SVs). The Transport Layer SVs are built by York Space Systems, Lockheed Martin, and Northrop Grumman.

These SVs are considered to be compact designs with total mass of 200-400kg (440-880 lbs). Deployed configurations feature body-mounted or single-hinge solar wings up to ~3-6 m² per panel (~ 30- 65 sq ft) for power and provide up to 3,500 watts of peak power. York Space Systems tranche 1 satellites leverage Hall-effect thrusters for propulsion. These electric propulsion systems operate on xenon or krypton gas to minimize propellant mass.​

PWSA – the Overall Strategy

The Proliferated Warfighter Space Architecture is composed of seven layers:

1. Transport Layer – Provides the resilient, low‑latency mesh network for data transport and tactical communications among satellites and to ground stations.

2. Tracking Layer – Conducts missile warning including detection and tracking of advanced and hypersonic threats using infrared and related sensors.

3. Custody Layer – Handles intelligence, surveillance, and reconnaissance (ISR) custody of ground and surface targets to support targeting and battle space awareness.

4. Battle Management Layer – Provides space‑based command and control, battle management, and mission planning/automation across the architecture.

5. Navigation Layer – Supplies alternative positioning, navigation, and timing (PNT) for satellites to complement or back up GPS based methods in degraded or denied environments.

6. Deterrence Layer – Monitors deep space to detect and characterize potentially hostile or suspicious actions against U.S. and allied space assets.

7. Support Layer – Encompasses ground segment, launch, and other infrastructure and services that enable operations of the space segment and other layers.

Transport Layer

The PWSA’s Transport Layer is a constellation of satellites that serve as the “connective tissue” for the entire architecture, utilizing optical inter-satellite laser links to move massive amounts of data across the network mesh in seconds. The two primary layers of the architecture, Tracking and Transport, will operate together as interconnected constellations of satellites in Low Earth Orbit. Tracking Layer satellites will detect and maintain custody of ballistic missile threats, while communicating their trajectory to the warfighter through a network of Transport Layer satellites in real-time.

The Transport Layer is a globally accessible mesh network known as Network Established Beyond the Upper Limits of the Atmosphere (NEBULA). NEBULA implements high-speed optical and radio frequency (RF) links on Transport Space Vehicles (satellites). NEBULA defines how data traffic is addressed, routed, secured, and managed over the mesh.

NEBULA is an implementation of a Multiprotocol Label Switching (MPLS) network, a well-established and standardized packet-forwarding technique in which routers move data traffic through a network based on short fixed-length labels instead of full network-layer addresses, creating predetermined label-switched paths (LSPs) that can provide predictable performance and quality of service.

Transport Space Vehicles (SVs) are equipped with:

• Optical Communications Terminal (OCTs) to connect SVs with other OCT equipped platforms

• Ka-band (RF) uplink/downlink for direct connect for mission data and command and control

• STTC – Space-Time Trellis Codings as back up for the primary Ka link

• Tactical Data Links (Link 16, 960–1,215 MHz TDMA) to ground based terminals on land, air, and maritime locations

• SBA network interfaces for IP-over-MPLS backbone per the NEBULA standards

• BMCs – Battle Management Command modules(s) for mission data processing

Optical Inter-Satellite Links (OISLs) will interconnect the Transport Layer Satellites with Optical Communications Terminals (OCTs).

Transport Layer Network Standards - OCT and OISL

Optical Intersatellite Link (OISL)Standards ensure interoperability across vendors and therefore are key to the implementation of PWSA. These standards enable high-speed, low latency near infrared laser communications between satellites in Low Earth Orbit. OISLs enable multi-gigabit (312.5 MHz to 2.5 GHz) connectivity with On-Off Keying modulation. OISL is key for supporting pointing, acquisition, and tracking (PAT) with spiral scans and finite-state machines for reliable acquisition of laser-based communications. PWSA integrates OISLs via Optical Communication Terminals (OCTs) to mesh the satellite constellation, providing resilient space-to-space links up to 6,500 km.

The SDA OCT Standard defines a common free‑space laser communications interface (Physical Layer + Error‑Correction and Frame‑Synchronization processing) so a variety of satellites and ground/air nodes can interoperate at multi‑Gb/s data rates.

The OCT standards provide specifications in the following domains:

Optical wavelength and modulation - definitions for physical properties for laser parameters such as wavelength, channel spacing, and spectral width to ensure interoperability with optical carriers. OCT standardizes a near‑IR wavelength region – typically in the 750–2500 nm wavelength range which is suitable for free‑space optical communications (FSOC) aka transmission across space. Near‑infrared refers to the part of the infrared spectrum just beyond visible red, typically from about 750 nm up to around 2500 nm.

Laser transmit power, irradiance, and beam shaping – The OCT Standard expresses optical performance predominantly in terms of received irradiance at a partner aperture, combined with beam size/divergence and BER (Bit Error Rate) requirements.

Irradiance requirements specify the required strength aka “how bright” the laser beam is at the receive point. For “short‑to‑medium” range links (hundreds of km), the OCT specifies an irradiance at the remote receive aperture of at least 25 µW/m² at a range of 100 km, while meeting specified error rates. For very long‑range links (thousands of km) the OCT specifies an irradiance ≥ 6 µW/m² at 20,000 km.

Expressed as power per unit area, irradiance is measured here in microwatts per square meter (µW/m²). So, in the OCT specification 25 µW/m² means that each square meter at the receiver plane gets at least 25 millionths of a watt of optical power.​ In OCT, the transmitted beam must deliver a minimum “brightness” or optical power density of 25 µW/m² or 6 µW/m² depending on the distance being transversed, despite beam spreading, pointing jitter, and other real-world effects.

OCT specifies optical beam divergence and spot size – The transmitted beam must have a full‑width at half‑maximum (FWHM) divergence, which is a measurement of the angular spread of the beam of no less than roughly 15 µrad. This i5 µrad specification in OCT means the laser beam spreads out very slowly, it is a very tight beam that only widens by a few millimeters per kilometer of travel, defined using the half‑maximum intensity criterion. Further, OCT specifies a beam spot size of at least about 1.5m diameter at a distance greater or equal to 100 km and at least ~0.5 m diameter at a distance of greater or equal to 500 km.

Receiver sensitivity, BER (error correction), and coding – These specifications establish baselines for error rates and sensitivity to ensure that if two terminals can see the same photons at the required signal-to-noise ratio, they can interpret each other’s frames and recover bits with the standardized error correction performance.

Pointing, acquisition, tracking (PAT) and link geometry – given OCT’s very tight beam requirements, PAT requirements are critical to ensuring successful operations because they govern how tight beams are aligned and held between moving platforms > aka satellites. Pointing control requirements ensure pointing stability and control to maintain a transmitted beam that meets the divergence and spot‑size requirements (see optical beam divergence above) at ranges ≥100 km and ≥500 km while achieving acceptable error rates.

PAT requirements apply to space‑to‑space (S2S) links, space‑to‑air and space‑to‑terrestrial links (S2T), so compliant OCTs can communicate with aircraft or ground optical terminals using the same physical and coding standards.

What are the exact laser wavelength and channel spacing requirements in SDA OCT

The current OCT specifications require a wavelength band based on the ITU‑T G.694.1 100 GHz DWDM grid. In other words, OCT uses the standard telecom C‑band, 1530–1565 nm, on the ITU 100 GHz grid, and is restricted to 44 channels.

Channel spacing is 100 GHz between adjacent channels, directly inheriting the ITU‑T G.694.1 100 GHz spacing. The channels are indexed such that each step in index n shifts the center frequency by 100 GHz; this allows wavelength‑division multiplexing of multiple SDA OCT channels on the same optical path.

Tracking Layer

The Tracking layer is a second constellation of more than 100 dedicated satellites, and it provides remote sensing and Earth observation via shortwave infrared (SWIR) sensors from LEO to spot, track, and support fire-control for advanced missile threats, including hypersonic missile systems. Tracking satellites are connected to the Transport Layer via the NEBULA optical mesh network with tracking data transmitted across the mesh network and down linked to the ground.

Fire‑control support will emphasize “targeting‑quality” and “stereo fire‑control‑quality” data products to integrate with air and missile defense systems. OISLs connect Tracking Layer satellites to each other and to Transport Layer Satellites, creating a resilient mesh network (NEBULA) for rapid dissemination of tracks and alerts.

“Wide Field of View” (WFOV) Overhead Persistent Infrared (OPIR) Payloads

The WFOV Overhead Persistent Infrared payload is designed to detect and track advanced missile threats from Low Earth Orbit, using large-format IR optics and on‑board processing. In general, these IR sensor arrays can be wide‑field “staring” sensors, which continuously image a large portion of the Earth, or scanning sensors that sweep their line of sight to build up a full scene.

The IR sensors used in PWSA satellite use a large‑format non‑scanning IR focal plane with many pixels, enabling persistent coverage over a broad angular sector with sufficient sensitivity to detect the intense heat from missile plumes and other hot objects against the colder Earth and space backgrounds. In other words, the IR sensors in PWSA are continuously watching a large portion of the earth’s disk. The hot exhaust plume of a boosting missile presents a bright IR source, which stands out spectrally and radiometrically, and the sensor measures radiance and converts it into digital signals representing brightness over its focal plane array

OPIR sensors typically operate in mid‑wave and/or long‑wave infrared bands tuned to missile plume and reentry signatures, supporting discrimination between threat objects and noise/clutter (e.g., clouds, sun glint, and earth’s terrain).

• Mid-wave infrared (MWIR) and Long-wave infrared (LWIR) are key bands in the infrared spectrum used for thermal imaging and sensing. MWIR spans 3–5 µm (micrometers), making it sensitive to moderate temperature emissions like engines or vehicles. LWIR covers 8–14 µm, ideal for detecting lower temperature sources such as human bodies or ambient environments.

• MWIR suits high end uses like surveillance, remote sensing, and target detection due to superior resolution. LWIR is common in night vision, industrial inspections, and security for its reliability in low-light conditions.

On‑board signal processing is real-time and performs detection and initial characterization of missile events. Onboard processing detects potential events in the raw IR imagery, performing initial filtering and thresholding to identify candidate missile launches or other significant heat sources.

PWSA Satellite Deployment s and Status – Tranche Methodology

To engineer and build a complex system like PWSA, while still adhering to its “move fast” mandate, the SDA broke the architecture down into bite-size chunks known as “tranches.” Tranches or iterations are designed to add additional capabilities with each generation deployed every two years – hopefully! Iterations of the architecture will likely integrate sophisticated quantum-resistant encryption and automated battle management directly into the laser-linked NEBULA mesh network.

Overview of Tranches

• Tranche 0 — Warfighter immersion: The minimum viable product is demonstrating the feasibility of the proliferated large satellite architecture in cost, schedule, and scalability.

• Tranche 1 — Initial warfighting capability: Regional persistence for tactical data links, advanced missile detection, and beyond line of sight targeting.

• Tranche 2 — Global persistence for all in Tranche 1. This will incorporate lessons learned from operating Gen 0 for at least two years.

• Tranche 3 — Advanced improvements over Tranche 2. This includes better sensitivity for missile tracking, better targeting capabilities and advances in blue/green laser communications and protected RF communications.

• Tranche 4 —Continual advances to the layers, including additional capabilities identified as current or future threats to the warfighter.

Deployment Status

The PWSA program has publicly described satellite launch schedules for Tranches 0 and 1 and has created planning markers around Tranches 2 and 3.

Tranche 0: Original plan = 28 operational space vehicles

• Launch 1: 10 of the Tranche 0 satellites occurred on April 2,2023 via Falcon 9 from Vandenberg SFB.

• Launch 2: 10 Tranche 0 satellites occurred on Sept 2,2023 via Falcon 9 from Vandenberg, one transport satellite remains on the ground as a software testbed.

Tranche 1: Original plan = 154 operational space vehicles plus 4 demonstration vehicles

• Launch 1: deployed 21 data transport satellites on Sept 10,2025 built by York Space Systems from Vandenberg Space Force Base.

• Launch 2: deployed 21 satellites on Oct 15, 2025, from Lockheed Martin from Vandenburg

• Additional missions are planned for 2026. No dates have been released.

Trance 2: Details for Tranche 2 launch timings have not been made public.

Tranch3: Dec. 19, 2025 - Space Development Agency announced the award of four contract, with a total value of approximately $3.5 billion, to build 72 Tracking Layer satellites in the PWSA program. The first of the satellites is expected to launch during the fourth quarter of fiscal 2029, according to the April solicitation – but this timeline has not been formally confirmed by SDA.

Challenges

Suffice to say the Proliferated Warfighter Space Architecture program is a complex major undertaking. White the program is well underway with deployments of Trance 1 satellites continuing in 2026, the program has experienced challenges stemming from supply chain disruptions, technical immaturity, and aggressive timelines.

• Delays in critical components like optical communications terminals, propulsion systems, and encryption devices have pushed back satellite readiness.

• Tranche 1 launches—originally set for Sept 2024 were pushed into late summer 2025 and now into 2026.

• Underwhelming Tranche 0 performance (as reported by the GOA) may cascade risks into Tranches 1-3

Wrapping up

PWSA is a strategic effort by the Space Development Agency. PWSA is transforming satellite to satellite and satellite to ground communications from legacy RF technologies to today’s mainstream optical systems, and at the same time implementing large constellations of small satellites in LEO. Both scenarios are highly sophisticated endeavors. PWSA is an interesting program that bears watching, because as we know, systems that are first brought to life by projects to support military operations, very often lead to spin off new implementations in the commercial space and to the creation of new supply chains and market segments.

Thanks for reading! Please, share your thoughts via the Comments section for this post or open a new SubStack chat thread … and please forward this post to your friends and colleagues. Until next time!

Beyond cryptocurrencies, ECDSA is also employed in secure communication protocols like Transport Layer Security (TLS) and Secure Shell (SSH). In these settings, ECDSA serves to authenticate the messages exchanged between clients and servers to ensure that the parties involved in the communication are who they claim to be and that the data transmitted has not been tampered with.

Overview of Elliptic Curve Digital Signature Algorithm

The ECDSA used by Bitcoin is based on a particular elliptic curve called secp256k1 and it plays a central role in securing crypto transactions by ensuring the security of crypto funds and verifying the authenticity of transactions.

ECDSA provides the foundations and mechanisms for users to sign their transactions with a private key, which is then verified by other nodes on the network using a corresponding public key. This process guarantees that only the rightful owner of the funds can initiate a transaction, and it also ensures that the transaction details, such as the amount and recipient, remain unaltered during the transfer.

What is an Elliptic Curve?

Elliptic curves are special form of a mathematical curve defined by an equation like y^2=x^3+ax+b. As mentioned above, Bitcoin, uses a specific curve called secp256k1 – see image below. In simple terms, an elliptic curve provides a one-way math function that is easy to use forward, but is practically impossible to reverse, ie. It is nearly impossible to crack the algorithm.

secp256k1 has gained popularity because its structure makes computation more efficient, often more than 30% faster than other curves, if the implementation is sufficiently optimized.

ECDSA in Action

ECDSA is used to create digital signatures that digitally “sign” crypto transactions, so that anyone can check that the transactions truly came from the owner of the coins, without learning the private key. In this way transactions are protected so they cannot be altered during transmission. The signing process looks like this – think of this as the “generic” process. Below we will outline more specifically how blockchain wallets execute transactions using a similar process.

• Generate a key pair - secret private key (only you know this key), and a public key that others can see.

• Build a crypto transaction and hash it.

• Produce a signature (r,s) for that hash – your wallet and ECDSA does this

• Broadcast the transaction plus signature to the network.

• Verify signature - recipient (and anyone else for that matter) uses your public key and the same hash to verify that the signature on the transaction is valid, proving you authorized the transaction. The transaction can then be executed.

Generate keys (once per wallet)

When you create a new crypto wallet, it does two main things with ECDSA. First it picks a large random secret number- this is your private key (which is never shared). Elliptic-curve math is then applied to compute a matching public key from the private key. Public keys are shared. Anyone can know your public key (or an address derived from it), but only you know the private key, and the underlying ECDSA math makes it practically impossible to go backwards from public key to compute the private key.

Create a transaction

To spend crypto coins, your wallet builds a transaction message that includes things like:

• Which coins/UTXOs you are spending.

• Who you are sending them to – a recipient address.

• How much you are sending, and any change back to yourself.

UTXOs are Unspent Transaction Outputs used by Bitcoin and other blockchains, and they represent distinct amounts of cryptocurrency received in a wallet but not yet spent – aka like cash in your pocket.

The wallet then “hashes” the transaction data through a cryptographic hash function (a mathematical function) that generates a fixed sized string of characters – called a hash value, or simply a “hash”. Bitcoin uses the SHA-256 hash.

Create the signature - Sign the transaction

Using your private key and the transaction hash, the wallet computes an ECDSA signature. In simple terms this computation looks like this:

• The wallet picks a fresh random number “k” for this signature only (never reused). This random number is known as a nonce and it is generated for each signature to ensure that each signature is unique and prevents predictable patterns.

• Elliptic-curve math is applied with “k” to compute a point on the elliptic curve and that is a number “r”. It combines “k”, the transaction hash, and your private key in a formula to compute another number called “s”. The signature is the pair (r,s), which gets attached to the transaction.

Anyone who sees the transaction sees the transaction data, your public key (or a form of it), and the signature (r,s), but still cannot recover your private key. The hash is what gets “signed” by ECDSA and this ensures that any change in the transaction changes the hash making the change detectable – aka if someone/something messes with your transaction, that can easily be detected, and the transaction can then be invalidated.

Send/broadcast the signed transaction

You are now ready to execute the transaction. Your wallet packages the transaction data plus the ECDSA signature (and usually the public key or a reference to it) and broadcasts this package to the cryptocurrency network (nodes and miners/validators).

Nodes that receive the transaction do not trust it blindly; they treat it as a “claim” that you authorized spending those coins, and they then check/verify the signature in the transaction to prove the transaction is legitimate.

Verify the signature

Each node performs ECDSA verification using the transaction data (to recompute the hash), the public key corresponding to the claimed sender, and the signature (r,s). Essentially the verification is simply confirming (or not) that the signature in the transaction matches the hash and the public key under the ECDSA rules.

If the verification passes, the node is convinced the transaction was made by someone using the correct private key (aka the public and private keys are a legitimate key pair). The node is also convinced that the transaction data was not changed after signing, because any change would alter the hash and that invalidates the signature. The transaction can then be executed onto the blockchain.

If verification fails, nodes reject the transaction as invalid, preventing the transaction from executing onto the blockchain.

Public-Private Key Pairs – The Keys (literally) to the Bitcoin Kingdom

Public-key cryptography is a form of asymmetric cryptography where two mathematically related keys play different roles vs. symmetric cryptography which involves the sharing of one secret key by both a sender and a receiver.

Public-key cryptography is a foundational technology and mechanism used to secure a vast array of systems including securing data transport over the Internet via HTTPS / TLS, securing email (PGP, S/MIME), digital signatures, time‑stamping and audit trails, software updates and code signing >> and of course securing blockchain powered digital cash, tokens, smart contracts, stablecoins, and NFTs etc.

Each key pair consists of a public key and a corresponding private key. The public keys can be disseminated widely and openly, and only the corresponding private keys need to be kept secret.

Key pairs are generated with cryptographic algorithms based on hard to solve mathematical problems termed one-way functions. A public‑private key pair is generated together by an algorithm (RSA, ECC, etc.) so that the keys are linked by a hard mathematical problem (e.g., factoring, elliptic‑curve discrete log). A private key is essentially a randomly generated number. A public key is a number that directly corresponds to a private key. A public key can be calculated from a private key, but not vice versa.

Weaknesses

As with all security-related systems, there are various potential weaknesses in public-key cryptography. Aside from making a poor choice of a key generating algorithm or selecting a key length too short, the greatest risk is that the private key gets exposed.

With the advent of quantum computing, many asymmetric key algorithms are considered vulnerable to attacks, and new quantum-resistant schemes are being developed to overcome the problem. Researchers and crypto experts the world over are working on new cryptography methods and algorithms to address this quantum risk. See this post on Post Quantum Cryptography.

How does Bitcoin leverage ECDSA? A quick overview of a Bitcoin Transaction

In very simple terms - a Bitcoin wallet is essentially a manager of key pairs and addresses. Wallets are designed to implement ECDSA as well as provide some extra bookkeeping and coordination capabilities.

Bitcoin uses ECDSA for ownership and authorization of spending. When someone pays you with Bitcoin, the output script “locks” the incoming coins to a condition involving your public key or address. Essentially it provides a signature that verifies against your public key. When you spend coins, your wallet builds a new transaction and provides an ECDSA signature (and usually the public key) that satisfies locking script conditions.

Wallets generate private keys with a cryptographically secure random generator and derive public keys and addresses from them. Many wallets use HD (hierarchical deterministic) schemes so many keys are derived from a single master seed, making backup easier.

In creating transactions, a wallet decides which UTXOs to spend, constructs a transaction, and hashes the relevant parts to get the message to sign. It signs with the corresponding private key using ECDSA, producing a signature and includes the public key as needed in the scriptSig script.

ScriptSig is the unlocking script in a Bitcoin transaction input. It provides the data (like signatures and public keys) needed to satisfy the locking script (ScriptPubKey) from a previous unspent transaction output (UTXO). ScriptSig works in tandem with the ScriptPubKey script to validate spending.

In simple terms, a Bitcoin wallet’s main job is to keep your private keys safe, construct correct transactions, and run ECDSA correctly for each input to verify and execute transactions.

Wrapping up

ECDSA provides secure and efficient digital signing and verification processes for crypto transactions. It has established itself as the standard for securing transactions in major blockchain networks – for now!

As briefly mentioned, the alarm bells are now ringing about the pending and serious risk posed by quantum computing and its potential for cracking existing crypto algorithms and related technologies, including ECDSA! Bottom line - blockchain networks will need to evolve, refine and adopt new cryptographic techniques to prevent being compromised by future quantum computing systems. The good news is that the crypto industry has recognized this challenge and the work to transition to new cryptography algorithms is underway.

Thanks for reading! Please, share your thoughts via the Comments section for this post or open a new SubStack chat thread … and please forward this post to your friends and colleagues. Until next time!

Let’s get started! Net-net, encryption must evolve to survive.

First – a Brief History Lesson

The word cryptography comes from the Greek words kryptos, meaning hidden, and graphien, meaning to write. Humans have been advancing “hidden writing” for thousands of years. Early evidence indicates symbol replacement hieroglyphics were used in 1900 BC Egypt and in 1500 BC, a Mesopotamian scribe was used to conceal a formula for pottery glaze.

To protect military messages from adversaries, the Romans (in the time of Julius Caesar ~ 100 BC) created the “Caesar Cipher”, a simple encryption method that shifted letters in the alphabet by a fixed number. For example, shifting each letter forward by three would turn “HELLO” into “KHOOR.” Only someone who knew the shift value (the key) could decipher the message. Variations of the Caesar Cipher were used for centuries, influencing the evolution of encryption into more complex forms.

Naturally, over time, the tools, techniques and encryption algorithms have evolved from manual systems into automated systems, first using dedicated electrical physical devices, and then into software and digital code. One of the most famous moments in encryption history, was the battle between Allied code breakers and the German Enigma machine during World War II.

This complex device used rotating cipher wheels to scramble messages, where each keystroke changed the encryption pattern, making every message unique creating nearly unbreakable communications for the Nazi military. Without the correct key settings, decrypting an intercepted message was nearly impossible. Led by British cryptanalyst Alan Turing and his team at Bletchley Park, the Allies famously cracked the Enigma machine coding algorithm, and this development gave the Allies a crucial advantage in the war.

First the scary/dire prediction part – the pending “Q-Day”

Fast forward to today, and the next evolution in cryptography - referred to as Post-Quantum Cryptography (PQC). If you have spent any time exploring quantum computing you have probably run across the term PQC, and the huge looming threat quantum computing poses for potentially breaking today’s core public key encryption methods like RSA. Instead of billions of years, it will be possible for a quantum computer to break RSA encryption in days or even hours, putting everything from state secrets to bank account information at risk.

Currently, many encryption algorithms rely on the difficulty conventional computers have with factoring large numbers. Today’s cryptographic algorithms select two very large prime numbers — which are only divisible by 1 and themselves — and multiply them to obtain an even larger number. While multiplying the prime numbers is easy and fast, it’s far more difficult and time-consuming to reverse the process and figure out which two prime numbers were multiplied together. These two numbers are known as the “prime factors.” For large enough numbers, a conventional computer has been estimated to need billions of years to reverse engineer the prime factors.

Simply stated, PQC is developing new cryptography methods that can withstand quantum attacks to prevent the so-called “Q-Day”. Q-Day is an estimate of the point in time when quantum computers will be able to reliably break existing RSA-2048 cryptography. The most common estimates put Q-Day in the early to mid-2030s, though timelines vary widely. NIST and NSA warn of possible arrival by 2030 with unforeseen breakthroughs in quantum computing. Bottom line, while we know Q-Day is coming, the exact date is unknown. Bottom line, current encryption algorithms will not be secure forever. So, researchers, crypto experts, governments and tech companies are already testing quantum-resistant alternatives.

The NIST PQC Program

A key player in the worlds of cryptography and quantum science is NIST – The National Institute of Standards and Technology in the U.S. NIST is a non-regulatory agency operating within the U.S. Department of Commerce. Its core mission is to promote U.S. innovation and industrial competitiveness by advancing measurement science, standards, and technology in ways that enhance economic security and improve quality of life.

NIST focuses on three main missions:

• Measurement Science (metrology) –- for precise units like time and voltage used in trade, commerce, product safety, and infrastructure

• Voluntary Standards and Guidelines – creating formal standards to be adopted globally (including for the US Federal Government and its agencies) for industries like information technology, cybersecurity and manufacturing

• Research in Emerging Fields – including quantum science, AI, and advanced manufacturing.

NIST receives primary funding through annual federal appropriations from Congress, typically around $1-1.3 billion. NIST is a world leader that has long played a key and highly valuable role in spurring research, development and implementation of standards for the most advanced cryptographic techniques and technologies.

PQC Program and Competition

Academic research on the potential impact of quantum computing on cryptography dates to at least 2001. A NIST published report from April 2016 cites experts that acknowledge the possibility of quantum technology’s potential to render the commonly used RSA algorithm insecure by 2030. (see Q-Day above)

As a result, NIST recognized the need to standardize quantum-secure cryptographic primitives and in December 2016 NIST initiated a new multi-stage standardization process by announcing a call for proposals to advance PQC. On August 13, 2024, NIST released final versions of its first three Post Quantum Crypto Standards.

Responses to the call for proposals attracted submissions for 23 signature schemes and 59 encryption/KEM schemes (public key–based schema) for the initial submission deadline at the end of 2017. A total of 69 proposals were deemed complete and proper and were moved into the first round of assessment/review. NIST engaged the world’s leading cryptograph researchers and experts to analyze and attempt to crack the candidate schemes which reduced the number of candidates. This entire process was open and transparent.

Candidates moving on to the second round of assessment/review were announced on January 30, 2019. On July 22, 2020, NIST announced 7 finalists (”first track”), as well as 8 alternate algorithms (”second track”). The first track contains the algorithms which appear to have the most promise and will be considered for standardization at the end of the third round. Algorithms in the second track could still become part of the standards, after the third round was completed.

First a word from our sponsor – NIST FIPS

One of NIST’s core mandated is the development, creation, and publication of Federal Information Processing Standards (FIPS), which are mandatory standards for U.S. federal systems. Given the high level of sophistication and diligence that goes into FIPS, in addition to the U.S. federal agencies, FIPS standards are adopted voluntarily by organizations around the world to support robust and secure information and data handling systems that are consistent and interoperable.

The first 3 PQC FIPS standards provide detailed descriptions of post-quantum encryption and digital signature algorithms so they can be implemented consistently to facilitate secure and interoperable communication. These new standards are designed for two essential tasks for which encryption is typically used:

• general encryption, used to protect information exchanged across a public network

• digital signatures, used for identity authentication.

Federal Information Processing Standard (FIPS) 203:

203 is intended as the primary standard for general encryption and it will be used to establish a shared secret key over open (aka insecure, like the Internet) channels. Among its advantages are comparatively small encryption keys that two parties can exchange easily, as well as their speed of operation. The standard is based on the CRYSTALS-Kyber algorithm, which has been renamed ML-KEM, short for Module-Lattice-Based Key-Encapsulation Mechanism.

Federal Information Processing Standard (FIPS) 204:

204 is intended as the primary standard for generating cryptographic keys, protecting and verifying digital signatures. The standard uses the CRYSTALS-Dilithium algorithm, which has been renamed ML-DSA, short for Module-Lattice-Based Digital Signature Algorithm.

Federal Information Processing Standard (FIPS) 205:

205 is also designed for digital signatures. The standard employs the SPHINCS+ algorithm, which has been renamed SLH-DSA, short for Stateless Hash-Based Digital Signature Algorithm. The standard is based on a different math approach than ML-DSA, and it is intended as a backup method in case ML-DSA proves vulnerable. Much like 204 this algorithm is well suited for remote digital signing, and as such can be thought of as a backup/alternative for 204.

Federal Information Processing Standard (FIPS) 206 = Work in Progress

Work on FIPS 206 is progressing, is currently in draft, and built around FALCON algorithm. 206 will be referred to as FN-DSA, short for FFT (fast-Fourier transform) over NTRU-Lattice-Based Digital Signature Algorithm. While there is no formal release date planned, estimates are 206 will be published in late 2026.

The three finished standards 203, 204, 205 are now ready for use, and organizations have already started integrating them into their information systems to future-proof them against the looming Q-Day “event”.

What the heck is “Lattice -Based” anything?

Post-quantum cryptography research is unfolding along several approaches and algorithm types. In this post we will not attempt to go deep into these approaches, but the primary approaches are:

• Lattice based – where a very large multi-dimensional grid of points, a lattice, is the mathematical foundation. Certain tasks on the grid appear very hard to calculate, such as finding the shortest non‑zero vector.

• Multi-variate – is based on the difficulty of solving systems of multivariate equations.

• Hash based – is a mathematical function that can map data of arbitrary size into fixed-size values. The values returned by a hash function are called hash values, or hash codes.

• Code based – leverages error-correcting codes where recovering a general codeword is made computationally difficult by applying transformations to a public key so it looks like a “random” code that is hard to decode.

• Isogeny-based – leverage the properties of isogeny graphs (group variables) of elliptic curves. Current Diffie–Hellman cryptographic mechanism uses this methodology.

As mentioned above, NIST standards 203 and 204 leverage the lattice-based schemes known as CRYSTALS-Kyber as the primary key-encapsulation mechanism (KEM), and CRYSTALS-Dilithium as a primary digital signature standard.

But Wait – there is more! FIPS 203 Back Up!

On March 11, 2025, NIST chose the Hamming Quasi-Cyclic (HQC) as the fifth algorithm for post-quantum asymmetric encryption and will be used for key encapsulation/exchange (KEM). This new algorithm will serve as a backup algorithm for ML-KEM (FIPS 203). HQC is a code-based scheme using different math than ML-KEM, thus mitigating possible weaknesses should any be found in 203. The draft FIPS standard for the HQC algorithm is expected in early 2026 with the final publication in 2027.

HQC and ML-KEM are both what experts call “key encapsulation mechanisms,” or KEMs. A KEM is used over a public network as a sort of first handshake between two parties that want to exchange confidential information.

Wrapping Up!

It is apparent that NIST and cryptography experts worldwide are making good progress on developing new super hardened cryptography algorithms and related standards to deal with the potential of Quantum computing cracking current cryptographic algorithms. This is critical work to support ALL organizations who use information (who doesn’t? ) 😊 to begin planning for the replacement of hardware, software, and services that use public-key algorithms so that data and information is protected from future quantum based attacks.

NIST has extensive experience developing encryption algorithms and the publishing of FIPS 203,204.205 is a first step on the journey to ensuring “Q-Day” is a non-event. With these new standards, and additional standards to come, technology managers can begin to inventory their systems for applications that use encryption, which will need to be replaced before cryptographically relevant quantum computers appear.

Let us know what you think. Please share your thoughts via the Comments section for this post or open a new SubStack chat thread … and please forward this post to your friends and colleagues. See you next time!

The U.S. Department of Defense is required by law to “maintain a Standard Positioning Service (as defined in the federal radio navigation plan and the standard positioning service signal specification) that will be available on a continuous, worldwide basis“ and “develop measures to prevent hostile use of GPS and its augmentations without unduly disrupting or degrading civilian uses”.

GPS is a set of radio technologies and related systems that falls within the larger category of Global Navigation Satellite System (GNSS). As of 2020, there are three fully operational GNSS systems:

• The U.S. navigation signal timing and ranging (NAVSTAR) GPS

• Russia’s Global Navigation Satellite System (GLONASS)

• Europe’s Galileo system

Though GPS technology is a subset of GNSS, receivers are differentiated as GPS only, or as GNSS. A GPS receiver is only capable of reading information from satellites in the GPS satellite constellation. A typical GNSS device can receive information from both GPS and GLONASS (or more than these two systems) at a time.

Because many of the global GNSS systems use similar frequencies and signals as GPS, around the L1 frequency spectrum, many “Multi-GNSS” receivers capable of using multiple systems have been produced.

The NAVSTAR GPS consists of 32 satellites owned by the U.S. and is the best-known and most widely utilized satellite positioning system. Russia’s GLONASS consists of 24 operational satellites with three used as spares and/or for use in testing.

GPS receivers—whether they are installed in ships at sea or embedded in a smartphone or a wristwatch—calculate their latitude, longitude and altitude by measuring the relative time delay of signals broadcast by at least four different satellites. Ground control systems, consisting of six monitoring stations, four ground antennas and Schriever’s master control station (MCS), communicate and operate the NAVSTAR satellites via communications from the ground antennas.

GPS Jamming - why do we need to care about it?

Because large swaths of the global economy, global navigation, aviation, and everyday basic security depend on GPS, the threat of disruption to GPS has major implications. It is very easy for crooks and other bad actors to leverage GPS jamming to support efforts in the drug trade, cargo theft and a host of other illicit activities. Drug traffickers regularly use jammers to try to foil electronic surveillance by law enforcement or rival gangs. The shipping security firm FreightWatch reported last year (2024) that there have been at least four large scale cargo thefts that were foiled in which GPS jamming devices were recovered. Simple jamming devices, despite being illegal to possess in the US, can be purchased for less than $200.00.

In 2013, the Federal Communications Commission fined a person almost $32k for using a device intended to evade the fleet management tracking system on his company vehicle. The device in question: a $30 GPS jammer.

For more than two years the FAA and New Jersey Port Authority were unable to determine why a new ground-based augmentation system (GBAS) – a flight system used primarily for augmenting aircraft take-off and landing at the Newark airport continued to experience intermittent failures. The cause of the failures was unknown. Leveraging specialized equipment, the culprit was finally identified. A contractor on site at the airport was using a GPS jammer that not only blocked his company vehicle’s fleet tracking system, it also took down the airport’s GBAS in the process.

A simple jamming device was able to take down a state-of-the-art, highly sophisticated aircraft landing system at one of the busiest airports in the world. Of note, this user was NOT trying to interfere with airport operations. Imagine what a person who DID intend to do harm to airport operations could do?

GPS is used for much more than just navigation. It’s also the primary source of timing and synchronization used in critical infrastructures such as financial, communications, industrial, the electric power grid, and more.

Jamming and Spoofing – so simple a child can do it!

Jamming is one challenge. Spoofing GPS devices is yet another challenge. GPS is susceptible to jamming and spoofing because its satellite signals arrive at the earth’s surface with extremely low power. GPS consists of three components: satellites, receivers and ground control stations (see GPS basics below). GPS spoofing and jamming are attacks that interfere with GPS signals. GPS jamming occurs when an external entity disrupts GPS signals by overwhelming them with noise or interference. This prevents receivers from determining an accurate location, leading to navigation failures, positioning errors, and loss of tracking.

Spoofing sends fake signals to trick GPS devices into showing incorrect locations, while GPS jammers block signals altogether, causing loss of service. GPS spoofing manipulates GPS signals to deceive receivers into believing they are in a different location.

Jammers broadcast noise on GPS frequencies, primarily L1 (1575 MHz) and L2 (1227 MHz), but at a higher power to drown out satellite signals. This makes it impossible for the GPS receiver to distinguish between the legitimate signal and the interference. The receiver is overwhelmed and fails to provide accurate positioning data. The three primary methods used in jamming include:

• Continuous Wave (CW): Concentrates full power on one frequency using a steady sinusoidal signal for targeted blocking.

• Sweep or Barrage: Often referred to as chirp jamming, these devices transmit across a band of frequencies and rapidly shift frequencies, typically between 1565-1585 MHz. In this way they can often overpower an entire range of signals rather than a single frequency.

• Directional: Uses antennas to aim interference at specific targets, such as drones or vehicles.

GPS jammers can be as simple as a device that plugs into a vehicle’s cigarette lighter port and are often marketed as anti-tracking tools despite legal restrictions. These types of devices are typically used to disable tracking in trucks and cars and broadcast on the L1 band at approximately 10mW.

Larger, complex, and more bulky devices with dozens of antennas, often referred to as hedgehog jammers (due to their many antennas) typically broadcast on the L1 or L2 band with around 10W of power. These jammers are often used by criminals seeking to jam multiple radio signals at the same time such as WIFI, GPS, and cellular signals and can cover kilometers – to disable alarm systems, hide the location of contraband or stolen materials, or prevent satellite-based navigation/tracking services.

NOTE: In many countries, including the United States, the use of GPS jammers by civilians is illegal. The US Federal Communications Commission (FCC) prohibits the sale, marketing, and use of any device that intentionally interferes with authorized radio communications because of the risk to public safety.

It is also possible for satellite malfunctions and/or solar flares to temporarily disrupt the transmission of GPS signals.

Consumer Facing Deployments

There are literally thousands of consumer and public applications and programs that rely on a robust secure GPS system. Some of them include:

• Civilian aviation/aircraft tracking relies on GPS.

• Global marine navigation relies on GPS.

• Disaster relief/emergency services - many emergency services depend upon GPS for location and timing capabilities.

• Geofencing/Geotracking use GPS to locate devices that are attached to or carried by a person, vehicle, or pet.

• Vehicle fleet operations - used to identify, locate and maintain contact with fleet vehicles in real-time for route optimization and operational efficiencies.

• Automobile navigation tools/software up to and including autonomous driving systems heavily leverage GPS.

• Satellite tracking and orbital operations including collision avoidance systems use GPS signals.

• Surveying/Tectonics/Mining - surveyors use absolute locations to make maps and determine property boundaries, and GPS enables direct fault motion measurement of earthquakes and to measure crustal motion and deformation.

• Telematics - where GPS technology is integrated with computers and mobile communications technology in automotive navigation systems.

Military Deployments

Military uses, not surprising, are also varied and diverse and include navigation over air/land/sea, radio clock synchronization, target tracking, missile and projectile guidance, satellite operations, and even nuclear detonation detection.

Fun Fact – GPS Break Through

GPS played a critical role in the 1991 Persian Gulf War, so much so, that this conflict has been characterized as was the worlds’ first “space war”. GPS supported U.S. and Coalition forces with navigation and position tracking. The first gulf war also demonstrated the susceptibility of GPS to being jammed. Iraqi forces installed jamming devices on likely targets that emitted radio noise, disrupting reception of the weak GPS signal.

More recently, in the current Russo-Ukrainian War, Ukrainian GPS-guided munitions have experienced significant failure rates resulting from Russian electronic warfare against GPS systems.

GPS Basics

The GPS project was launched in the United States in 1973 to overcome limitations of previous navigation systems. The U.S. Department of Defense developed the system, which originally used 24 military satellites, and it became fully operational in 1993. Civilian use (with limited capabilities) went operational in the mid-1980s.

The work of Gladys West on the creation of the mathematical geodetic earth model is credited as instrumental in the creation of computational techniques for detecting satellite positions with the precision needed for GPS. Mrs. West started her career at the Naval Surface Warfare Center in Virginia in 1956. At the time, she was second black women working for the US Navy.

How does it work?

GPS is a navigation system using satellites, a receiver and algorithms to synchronize location, velocity and time data for air, sea and land travel. It is owned by the United States Space Force and operated by Mission Delta 31 (MD31). MD31 is the United States Space Force unit responsible for navigation warfare.

GPS does not require the user to transmit any data, and it operates independently of any telephone or Internet reception, though these technologies can enhance the usefulness of GPS positioning information. While the U.S. government created, controls, operates, and maintains the GPS system, it is freely accessible to anyone with a GPS receiver.

A GPS tracker receives radio frequency (RF) signals in the microwave band from an array of satellite transmitters orbiting the earth. Once a GPS tracker receives signals from four or more satellites, it determines its position through a series of time calculations and trilateration. The receiver relies on these precise and specific satellite signals to determine where it is in the world.

System and Components

GPS is made up of three different components, called segments, that work together to provide location information. The three segments of GPS are:

• Space (satellites): The satellite system is a constellation of at least 31 satellites in six earth-centered orbital planes, each with four satellites, orbiting at 13,000 miles (20,000 km) above earth and traveling at a speed of 8,700 mph (14,000 km/h). The satellites transmit signals to users for geographical position and time of day. While only three satellites are needed to accurately resolve a location on earth’s surface, a fourth satellite is often used to correct the receiver’s clock error, to provide more accurate positioning.

• Ground Control: The Control Segment is made up of earth-based monitoring stations, master control stations and ground antenna. Control activities include tracking and operating the satellites in space and monitoring transmissions. There are monitoring stations on almost every continent in the world, including North and South America, Africa, Europe, Asia and Australia.

• User Equipment: Includes GPS receivers and transmitters, including items like watches, smartphones and telematic devices. In fleet operations, devices like telematics units or driver smartphones receive satellite signals and calculate exact positions to support dispatching and fleet safety. GPS receivers can be standalone/handheld devices for consumer use, or compact integrated circuits (ICs) or modules designed to be embedded into larger systems, devices, such as vehicles, sea going vessels etc.

As of 2025, 83 Global Positioning System navigation satellites have been built: 31 are launched and operational, 3 are in reserve or testing, 44 are retired, and 2 were lost during launch. GPS Satellite launches are on-going, as satellites need replacement after their useful lifespan. The next GPS satellite launch (GPS III SV0) is planned for late Jan 2026 from Cape Canaveral Space Force Station’s SLC-40 (launch pad) on a SpaceX Falcon 9 rocket.

The GPS satellite constellation requires a minimum of 24 operational satellites and allows for up to 32; typically, 31 are operational at any one time. After being launched, GPS satellites enter a period of testing before their signals are set to “Healthy”. During normal operations, certain signals may be set to “Unhealthy” to accommodate updates or testing.

After decommissioning, most GPS satellites become on-orbit spares and may be recommissioned if needed. Permanently retired satellites are sent to a higher, less congested disposal orbit where their fuel is vented, batteries are intentionally depleted and communication is switched off.

GPS uses three main civil radio frequency (RF) signal bands today: L1 at 1575.42 MHz, L2 at 1227.60 MHz, and L5 at 1176.45 MHz, each with different roles and signal structures.

• L1 band - L1 is centered at 1575.42 MHz and is the original and most widely used GPS civil band. It carries the open C/A code, the encrypted P(Y) code, and newer signals such as L1C and M‑code, supporting legacy civil, modern civil, and military users.

  The C/A code repeats every 1 millisecond and is used by civilian receivers to acquire and track satellites quickly. Each GPS satellite transmits a unique C/A code from a set of Gold codes, by modulating the code onto the L1 carrier frequency.

  The P(Y) code is an encrypted code that repeats every 7 days and is used primarily for military precise positioning service (PPS) on both L1 and L2 carrier frequencies. Only authorized receivers with cryptographic keys can decrypt P(Y), providing higher precision and jam/spoof resistance.

• L2 band -L2 is centered at 1227.60 MHz and was originally introduced mainly for military and precise applications. It carries P(Y) and the modern civil L2C signal, enabling dual‑frequency ionospheric correction (error correction for atmospheric affects) and higher‑precision positioning when combined with L1.

• L5 band - L5 is centered at 1176.45 MHz, in the Aeronautical Radionavigation Service band reserved for safety‑of‑life operations. L5 has higher transmitted power, wider bandwidth, and longer codes than L1 C/A, improving multi-path rejection, interference resilience, and accuracy for applications like aviation approaches.

Jamming Counter Measures

Anti-jamming techniques use several technologies to detect, isolate, and reduce the impact of GPS interference. Solutions may be hardware-based, software-defined, or a hybrid of both. The common methods used to counter/prevent GPS jamming can be categorized as follows:

Antennas and related tools: Multi-element high quality antennas can help counter GPS jamming through specialized designs that detect and suppress interference while preserving weak satellite signals. These systems create directional “nulls” toward jammers, enabling reliable navigation in hostile environments. Examples are:

• Controlled Reception Pattern Antennas (CRPAs) use multiple (4-16) antenna elements in an array to adaptively steer reception patterns. They generate nulls in the direction of jamming sources, canceling noise like covering an ear in a loud crowd, while focusing on legitimate GPS signals. Digital beam forming enhances this by processing signals to precisely nullify interference and boost signal-to-noise ratio.

  What is a “null”? A null is “zone of silence” in the direction of the detected interference. This suppresses the jamming signal while still allowing legitimate satellite signals from other directions to be received. Null steering is effective in many scenarios but has limitations: it typically works best against a single interference source and cannot address all threats, such as spoofing or complex multi-directional jamming.

• High-gain directional antennas: Focus on skyward GPS signals, rejecting ground-based jammers via narrow beam width and polarization matching.

• Active anti-jam antennas: Embed filters like a Surface Acoustic Wave (SAW) filter to selectively pass GPS signals while rejecting unwanted interference, or ceramic to block out-of-band interference.

Signal-processing algorithms: Signal-processing algorithms help GPS receivers fight jamming by spotting and removing unwanted interference, much like tuning out loud static on a radio to hear a faint voice. These methods analyze the incoming signal to separate the weak GPS data from stronger jamming noise.

Algorithms first detect jamming through methods like energy detection, which scans for unusual power spikes, or cyclostationary analysis (to characterize signal properties), which identifies repeating interference patterns. Once spotted, adaptive notch filters create a null in the signal at the jamming frequency, blocking it without harming the GPS signal. For swept (chirp) jamming that shifts frequencies, high-resolution tools like Fast Orthogonal Search model the interference precisely so again it can be mulled out. Combined with hardware like directional antennas (see CRPAs above), with these systems, receivers stay operational where standard ones fail.

Integration with Inertial Navigation Systems: GPS receivers can be combined/ integrated with other Inertial Navigation Systems (INS) to provide resilient navigation during jamming. INS uses gyroscopes and accelerometers to track position, velocity, and orientation independently of GPS signals, so can bridge outages caused by interference. INS essentially provides a backup mechanism if the core GPS system becomes overwhelmed by jamming.

Wrapping up

Not trying to be a “Dougie Downer”, but here is a sobering thought….

There are no formal plans for a full replacement of GPS. The U.S. Space Force key efforts are focused on refinement and augmentation by enhancing satellites, ground control, and user equipment to improve accuracy, anti-jamming resilience, and signal capabilities. GPS III Satellites (generation III) are currently being deployed, and they will deliver 3X greater accuracy and 8X better anti-jamming. The GPS IIIF program, with launches starting in 2027, will deploy up to 22 satellites with 60X greater anti-jamming.

Please, share your thoughts via the Comments section for this post or open a new SubStack chat thread … and please forward this post to your friends and colleagues. Until next time!

In the lingo of space flight operations, the probability of a collision is called a conjunction - a predicted close approach between two objects in orbit. A conjunction is not a collision itself, but a warning that two objects are expected to pass near each other closer than a preset safety distance. Satellite flight dynamics and operations teams use conjunction predictions to assess the probability of collisions and, if that probability is too high, plan and execute a small orbit change (a maneuver to alter the satellite’s Δ v, delta v, velocity) to keep the satellite safe.

Satellite Collision Avoidance Systems - Core Components

Current automated satellite collision avoidance, or conjunction assessment (CA), systems employ a three-stage process of screening, probabilistic risk assessment and autonomous maneuver planning or “mitigation”.

• Phase 1: Screening

  Spacecraft operators supply spacecraft orbit data, called ephemerides, to the USA Space Defense Squadron where the data is ingested and compared to a master catalog of all objects in orbit This catalog is maintained by the U.S. Space Force. Ephemerides are detailed descriptions of where a spacecraft is in space at different times, and how fast it is moving. Think of it as the “flight schedule and track” of a spacecraft. The orbit data is screened for close approaches and used to compute collision probabilities.

• Phase 2: Risk Assessment

  The screening data is then packaged into standardized files called Conjunction Data Messages (CDMs) that describe predicted close approaches and satellites state covariances. Covariances are estimates of errors in a satellite’s position and velocity and has three values for position (x, y, z) and three values for velocity that are presented as a matrix.

  The CDMs are then sent to the spacecraft operator who is responsible for analyzing the data, performing risk analysis, and executing collision avoidance (mitigation) steps if needed. Most space operators compute a “Probability of Collision” or Pc with a chosen threshold of action to assess if mitigation is warranted. Pc is computed using sophisticated analytical and data sampling methodologies and is used to model “virtual” encounters as a core aspect of the analysis.

  As you can imagine this conjunction/risk assessment is a highly specialized and complex undertaking. While commercial space operators have developed and operate in-house risk assessment capabilities, there are several government sponsored and commercial third party providers who offer specialized services. Examples include NASA’s Conjunction Analysis Risk Assessment (CARA) program, NASA Johnson Space Center Flight Operations Directorate, and the European Union Space Surveillance and Tracking (EUSST) program.

• Phase 3: Mitigation

  If a Pc or object approach distance exceeds operator thresholds, mitigation options are evaluated and a plan is chosen. To assess options, an automated planner searches candidate maneuvers (time, direction, magnitude) to minimize risk while preserving mission and flight objectives. Maneuvers must be screened and approved before execution to minimize the potential for creating additional close approaches. Approved maneuvers are uploaded or autonomously commanded, then after completed, the system rescreens using updated spacecraft position states to re-assess conjunction risks and Pc.

While the growing number of avoidance maneuvers is designed to make orbital operations safe, maneuvers can also have negative effects on predictions of future collisions. A study by the Pennsylvania-based Commercial Space Operations Center (COMSPOC), published last year, found that every collision-avoidance maneuver throws off satellite path forecasts for several days. In the aftermath of maneuvers, satellites’ actual positions may differ from their forecasted ones by up to 25 miles (40 kilometers), potentially making collision predictions inaccurate.

Autonomous onboard collision avoidance

Operators of large-scale satellite constellations (like SpaceX Starlink) deploy an automated collision avoidance system that can plan and execute maneuvers on-board every spacecraft. Here is an overview of the typical core components and capabilities of the multi‑tiered software architecture that underlies these collision avoidance systems. (these details are typical and are not necessarily specific to Starlink)

• Navigation and state estimation – an onboard navigator module (usually a Kalman filter) combines data from GPS and other sensors to produce precise estimates of position, velocity, and attitude in real time – aka the “state vector” of the spacecraft. The state data also includes estimates of possible errors or “covariance”. The spacecraft state is fed into the conjunction logic in real time.

• Conjunction logic, catalog and external inputs – conjunction logic are the rules and algorithms used to detect, evaluate, and respond to close approaches between a spacecraft and other objects. An External Catalog Application (or equivalent) maintains an onboard catalog of high‑risk objects derived from ground catalogs and CDMs and is on-board catalog is updated periodically.

• Autopilot, guidance and control – an autopilot application performs autonomous conjunction assessment and maneuver analysis to turn candidate maneuvers into feasible thrust sequences. Guidance and control software enforce mission and flight constraints (trajectories, attitude pointing, formation geometry for constellations, etc.) while minimizing Pc and future conjunctions.

• Supervisor/commander and mission manager – a supervisor or mission‑manager layer coordinates between the autopilot, navigation, and catalog applications. It triggers Pc checks at configurable times and handles go/no‑go maneuver decisions and timelines. The mission manager also coordinates with ground based Space Traffic Management (STM hubs) that synchronize multiple catalogs, planned maneuvers, and risk estimates across multiple operators to avoid conflicting autonomous responses. STM hubs manage fleets of satellites vs. managing individual satellites.

Ground based support systems - AI/ML technologies

Onboard autonomous systems are impressive for sure, and they require numerous support systems that are ground based. Here are few examples:

High‑volume screening and policy enforcement usually remain ground‑centric. Operators of large fleets of satellites typically run continuous screening on their fleets against global catalogs, available CDMs and send maneuver recommendations up to spacecraft. These processes involve scalable orbit propagation and management, covariance management (errors in a satellite’s estimated position and velocity), and prioritization (triaging) thousands of daily close call alerts down to a manageable number.

Coordinating with NASA’s Conjunction Assessment Risk Analysis (CARA) program is an example of these activities. NASA CARA predicts and evaluates close approaches between NASA’s uncrewed spacecraft and other space objects, then helps plan collision-avoidance maneuvers when needed, and is tasked to assist satellite operators in designing and executing mitigation maneuvers.

AI/ ML for criticality and maneuver recommendation. Systems like the European Space Agency’s AUTOCA (Autonomous Collision Avoidance) system is an AI‑driven decision support tool that ingests conjunction data messages for a satellite, and leverage AI/ML models to estimate conjunction criticality and predict Pc evolution over time and autonomously recommends if and how to perform a collision avoidance maneuver, particularly for large satellite fleets. The AI/ ML components operate on data extracted from CDMs (relative geometry, covariances, object class) and feed into rule‑based logic that compares predicted risk to operator policy thresholds.

How are maneuver Go/No Go decisions made?

Automated collision avoidance sounds like a no brainer – right? Deciding if/when/how to maneuver a satellite is not as straight forward as you might think. Ultimately deciding to maneuver is based on a bone fide quantified risk where important metric(s), usually a high probability of a collision Pc (geometric miss distance), crosses predefined thresholds within a specified time window, subject to spacecraft health, mission, and operational constraints. Assessing the various constraints is a complex multi-variate analysis. Further, any maneuver “burn” must be pre-approved and carefully planned to ensure the maneuver is optimized and does not result in additional Pc.

Bear in mind that the more maneuvers a satellite makes the faster it use up its thrust propellant, resulting in a shortening of its operational life, potentially leading to an earlier “de-orbiting” than is desired.

Here is a simplified overview of the core criteria used to decide to maneuver…

Probability of collision and miss distance – software computes probability of collision, Pc, using the current satellite state estimate and covariance for both objects with a maneuver considered if Pc exceeds a set limit and/or the predicted miss distance is below safety buffers.

Satellite operators are on the hook to establish and enforce these limits to ensure satellite operations are safe and secure. Pc limits are often time‑dependent (aka the shorter the time window to “closet approach”, the stricter the limit) and usually vary by object type: crewed satellite, space station, high‑value spacecraft, or random space debris.

An important criterion is Time to Closest Approach (TCA) window. Automation typically defines “decision windows” before TCA occurs and only considers maneuvers if a feasible burn can be executed successfully and settled before that time. Another way to think of this is that an assessment of the “last‑safe‑burn time” is needed. If the system is inside the TCA window without a viable solution, rules usually fall back to “do not maneuver” to avoid making things worse with late, or poorly orchestrated burns.

Navigation state and continuous assessment of uncertainty – onboard automation maintains a best‑estimate orbit plus covariance (error/uncertainty). High uncertainty can both inflate Pc and temporarily suppress maneuvers until more tracking data reduces ambiguity and uncertainty. Some implementations use “consider” parameters or covariance inflation when recent maneuvers or limited tracking make predictions less reliable, which directly affects the maneuver/no‑maneuver decision.

Conjunction data and context – as described above, the onboard software receives or maintains a list of candidate conjunctions with data from CDMs (from uplinked messages or onboard propagation) and periodically recomputes risk for conjunction. It also checks metadata such as whether the other object is maneuverable, whether there are multiple close approaches in a cluster, and whether prior planned maneuvers already mitigate the event under consideration.

Decision logic structure – rule‑based “gates” – most space flight implementations use a layered rule set: first, filter out obviously benign events; then apply numeric thresholds on Pc, both radial and along‑track separation, and timing; finally, apply mission and safety constraints before authorizing a maneuver. Radial separation refers to the difference in distance from the center of earth between two spacecraft – which spacecraft is higher or lower relative to the other. Along-track separation refers to the distance between spacecraft measured forward or backward along their flight path, aka who is ahead or behind the other and by what distance.

Safety constraints/rules that are typically considered include:

• enforce minimum separation (like commercial aircraft) in radial and along‑track directions

• spacecraft fuel and engine duty‑cycle limits – is there enough fuel and can the engine execute the burn effectively

• enforce “no‑burn” periods, such as during critical payload operations or pre-designated safe‑modes

Optimization/assessment of candidate burns – if a conjunction passes the rule gates, the automation generates several candidate maneuvers (varying time, direction, and magnitude) and scores them against a cost function combining residual Pc, and Δv, the desired change in a spacecraft’s velocity to perform a maneuver. Δv changes to spacecraft velocity is one of the core design and navigation parameters for any mission, because it directly links maneuver capability, propellant mass, and mission feasibility. Navigation burns are also assessed against the impact on future operations, and potential creation of new conjunctions.

The system then selects a solution that drives Pc below a lower “clearance” threshold while staying within Δv and operational limits. If no solution candidate satisfies these constraints, the automation may choose to stand down on executing a maneuver and monitor status only.

Constraints for safety, mission, and human‑in‑the‑loop considerations

A variety of spacecraft systems’ health and configuration checks are executed before any burn operation. Confirmations of propulsion health, attitude margins, and that required sensors and actuators are fully operational are completed. If any of these systems fail a check, usually the maneuver is blocked and flagged for human review. Some systems enforce “hold‑downs” after recent anomalies or large burns, meaning the automation can compute but not execute new maneuvers without human operator approval.

Human operator policies and coordination

Best‑practices for satellite operations recommend that even autonomous systems follow operator‑defined policies such as always communicate planned maneuvers, avoid simultaneous autonomous responses by multiple spacecraft, and respect agreed to “right‑of‑way” conventions when other operators are involved. Many current autonomous architectures let onboard software decide and stage a maneuver but leave a short veto window for ground personnel to approve/deny a maneuver, unless a rapid response is required.

What about situations when automation decides not to maneuver

As Pc gets updated regularly, if it shows improving geometry where the Pc is trending downward or dominated by catalog uncertainty, the logic may deliberately wait for more data rather than maneuver prematurely. At very high relative speeds or grazing geometries (the spacecraft’s path has a very shallow angle relative to another object) small burns may not materially reduce risk, so the automation may conclude that no maneuver is the safest option.

Automation logic will penalize maneuvers that push the satellite into neighboring planes or orbital areas where conjunction rates are higher, even if the maneuver would fix the immediate event. If a burn would consume disproportionate Δv relative to the spacecraft’s remaining lifetime, policies can force the system to accept a lower Pc risk instead.

Which onboard sensors trigger a maneuver decision

For satellite collision avoidance, no single “proximity sensor” like radar or lidar normally triggers the maneuver decision. Triggering maneuvers involve navigation and tracking data from a small set of core onboard sensors plus uplinked space object catalog data.

Primary onboard sensors can include GNSS receivers, GPS, and/or Galileo that track and provide the satellite’s precise position and velocity, which the onboard navigation filter turns into an orbit state and covariance.

A GNSS (Global Navigation Satellite System) receiver is a specialized electronic device that “listens” to navigation signals from multiple of position/location systems like GPS, Galileo, GLONASS, and BeiDou. GNSS times signals from several satellites, to calculate their own position, speed, and very accurate time. GNSS units have large, high‑quality antennas and low‑noise electronics, giving cleaner signals and less error from reflections and interference than systems like GPS. GNSS receivers can routinely reach sub‑meter to centimeter‑level accuracy when combined with error correction.

GPS is the US satellite navigation system used to support a multitude of military, commercial aviation, autonomous vehicles, mobility (phone) and consumer products and services. GPS satellites broadcast timed radio signals that let compatible receivers figure out their position anywhere on or near Earth. Accuracy of GPS systems in space is typically 10 meters horizontally and 20 meters vertically in low Earth orbit (LEO) but can be enhanced with specific optimizations. GPS is useful primarily for LEO only, as GPS signals weaken significantly beyond 1,800 miles.

Galileo is Europe’s satellite navigation system, similar to GPS but designed for very high accuracy and civilian control.

Inertial Measurement Unit (IMU), gyros and accelerometers measure attitude rates and non‑gravitational accelerations (from thrusters, atmospheric drag), improving short‑term knowledge of how the orbit is changing. Accurate modeling of these variables/changes tightens the uncertainty (covariance), which directly affects computed collision probability and therefore whether a maneuver is triggered.

Attitude‑determination sensors including star trackers, sun sensors, magnetometers capture the spacecraft’s orientation so the system can correctly project thrust directions and non‑gravitational forces into the orbit frame. Developing accurate attitude knowledge ensures that the predicted effect of any candidate maneuver on future conjunctions is reliable, which is necessary before autonomously authorizing a burn.

Finally, system health and configuration monitor telemetry from the propulsion system (tank pressure, valve status), satellite power system, and attitude‑control sensors also acts as a gate, meaning if they report out‑of‑limits conditions, or a malfunction, the automation will block a maneuver even if the conjunction risk is high.

Satellite operators are responsible for defining sensor thresholds that initiate onboard maneuvering and defining numeric limits on navigation quality and risk metrics, and these are what the onboard autonomy watches. Quality and risk metrics include GPS/GNSS estimation quality such as maximum allowable position error, velocity error, or covariance trace before an autonomous maneuver is allowed.

Operators tune sensor threshold values depending on satellite type and capabilities, mission objectives, acceptable risk, propulsion capability, and satellite constellation density. Typically, these values are not published publicly. Even within a satellite fleet, thresholds can differ between high‑value spacecraft, short‑lived smaller satellites, and deorbiting vehicles.

How does a satellite execute a burn to alter its position in space

Changing the orbit of a spacecraft is accomplished with an orbital maneuver called a “burn”. A satellite executes a burn by firing its thrusters in a precisely timed direction to change its velocity, or delta-v (Δv), which modifies its orbit. Even small changes in speed or direction can significantly reshape a craft’s orbit because orbital paths are set by the satellite’s position and velocity vector around the central body – aka usually the earth.

Space flight operations teams, or as discussed in this piece autonomous flight systems, determine and compute the exact Δv vector (magnitude, direction, and timing) needed to get from the current orbit to the target orbit. This typically involves choosing a maneuver point such as perigee, apogee, or a node where burns are most efficient for raising/lowering altitude or changing inclination.

There are three basic ways a spacecraft can apply thrust by firing its engine(s) to change its orbit: prograde, retrograde, and out‑of‑plane burns. These describe direction of thrust relative to the current path.

• Prograde fires the engine in the same direction the spacecraft is already moving along its orbit, akin to stepping on the gas pedal in a car, to speed the spacecraft up and raises the opposite side of the orbit

• Retrograde fires the engine opposite the direction of exiting motion, akin to applying the brakes in a car, it slows the spacecraft down and lowers the opposite side of the orbit

• Normal/Out‑of‑plane thrust is pointed above or below the orbital plane instead of forward or backward along the path. This changes the tilt (inclination) of the orbit, like tipping a hula hoop to a new angle in space, without raising or lowering the average height of the orbit.

At the planned time, the spacecraft propulsion system opens valves and ignites the thruster(s) (chemical, electric, or cold‑gas), maintaining attitude control so thrust stays aligned with the commanded vector for the duration of the burn. Once the required burn duration or Δv is reached, the thrusters shut off, leaving the satellite coasting in its new orbit defined by the updated position and velocity. Ground controllers and on-board navigation then carefully track the spacecraft, estimate the achieved orbit, and, if needed, schedule small cleanup burns (trim maneuvers) to fine‑tune its position.

Types of spacecraft propulsion systems

Satellite thrusters (on-board maneuvering engines) fall into a few major families that trade off thrust level, efficiency, complexity, and propellant type.

• Chemical thrusters create force by burning propellant and expelling hot gas at high speed through a nozzle; they can be monopropellant – aka single fluid design usually with hydrazine fuel, or “green” alternatives (usually nitrous oxide-based), or bipropellant, fuel plus oxidizer, systems. Chemical thrusters deliver relatively high thrust and are common for orbit raising, major attitude changes, and fast maneuvers, but their efficiency is modest, so they consume more propellant mass for a given total Δv.

• Cold‑gas systems store an inert gas, often nitrogen or sometimes compressed air or similar benign gas, in a tank and release it through a simple valve and nozzle without combustion. These systems are very simple, clean, and highly reliable, and are often used for small satellites and for attitude control—but they provide low levels of thrust, meaning they quickly run out of useful Δv for larger/long missions.

• Electric thrusters use electrical power (from solar arrays or batteries) to accelerate propellant ions or plasma to very high exhaust velocities, achieving much higher specific impulse than chemical or cold‑gas systems. Common electric types include ion engines and Hall‑effect thrusters, which offer extremely efficient long‑duration thrust ideal for station‑keeping, slow orbit raising, and deep‑space cruising. The downside of these systems is their need for substantial onboard power.

*Note: Nuclear - NASA has been exploring the potential for Nuclear Thermal Propulsion since the 1060’s. While nuclear systems are viewed as potentially viable and powerful these remain experimental technologies with NASA continuing to work towards an in-orbit demonstration….

Wrapping up

Operating large constellations of satellites would not be possible without sophisticated conjunction management systems powered by autonomous systems. Collision avoidance starts with establishing awareness of where everything is in space and where it is headed aka Space Situational Awareness. Hopefully this post helps illuminate the core processes and technologies being deployed to keep spacecraft where they belong – coasting safely in their designated orbits around the earth.

Please share your thoughts via the Comments section for this post or open a new SubStack chat thread … and please forward this post to your friends and colleagues. Until next time!

In early Dec, a just launched Chinese satellite came within 200 meters of slamming into an orbiting SpaceX satellite at 560 km altitude. Several additional serious near‑misses over the past year got me curious about what if any global satellite traffic‑management (STM) process or system is in place – or is it being actively worked on?

There are currently ~ 46,600 objects in orbit that are actively tracked, over 10,500 of which are active satellites and/or spacecraft transporting humans. These numbers do NOT include space junk. It is estimated that there are 50-55,000 objects larger than 10 cm (trackable with ground radar) and hundreds of millions of objects sized 1 cm to 10 cm, aka bits and pieces large enough to seriously damage or destroy a satellite on impact.

About 80 percent of operational spacecraft are in low Earth orbit (LEO)—that region of space within 200-2000 km above the surface of the Earth. SpaceX Starlink is currently operating ~ 8,300 satellites— roughly two-thirds of all the active satellites in orbit. China Satellite Network (China SatNet) is initially planning to launch ~ 6,000 satellites for a communications constellation with a longer term goal of having ~ 12,000 satellites orbiting. Amazon and Eutelsat OneWeb together will soon have an additional ~ 1,200 satellites floating above the earth.

Ten nations currently have domestic launch capabilities, with more countries rapidly pursuing their own launch capacity. Nearly 100 nations operate or use satellite data from spacecraft that are owned by national enterprises.

What is going on - recent near miss incidents

• As mentioned in the intro, in December 2025, one of nine satellites launched on a Chinese Kinetica‑1 rocket came within approximately 200 meters of a Starlink satellite at about 560 km altitude. SpaceX publicly highlighted that, as far as it knew, no prior coordination or deconfliction (collision avoidance) had been performed.

• In late August 2019, a weather satellite belonging to the European Space Agency (ESA) was forced to fire thrusters to avoid colliding with a SpaceX satellite.

• Europe’s OneWeb is currently operating about 660 low earth orbit satellites, and they have had multiple instances of collision warnings with SpaceX satellites.

• In February 2024, a 35-year-old Russian spy satellite and a 25-year-old NASA scientific satellite nearly collided, missing each other by just 33 feet.

• Dec 2024–May 2025, reports indicate Starlink satellites executed 144,404 collision‑avoidance maneuvers illustrating how collision alerts and autonomous deconfliction have become routine, even though satellites from many other operators lack comparable automated capabilities.

Tracking and Collision Avoidance Systems – Otherwise Known as Space Situational Awareness (SSA)

Collision avoidance starts with establishing awareness of where everything is and where is headed – in space – aka Space Situational Awareness (SSA). In short, SSA is the knowledge, characterization, and practice of tracking space objects and their operational environment. SSA data is used to predict conjunctions between objects and warn space operators of potentially dangerous close approaches to enable collision avoidance maneuvers. SSA is foundational to all space safety and real time space traffic coordination activities.

SSA services are more than just satellite tracking. SSA also includes space surveillance, satellite overflight, orbital awareness, object cataloguing, debris analysis, conjunction analysis, satellite integrity monitoring, coordination and communications, and many other activities. SSA enables space actors to engage in decision-making and to perform actions based on tangible in-orbit related intelligence.

Who is responsible for SSA?

The primary satellite tracking and SSA providers today are a mix of national military systems, notably the US Space Force’s 18th Space Defense Squadron (18 SDS U.S.), commercial space operators (SpaceX, BlueOrigin), US Defense Contractor Primes (LockHeed Martin, Northrop Grumman), NASA, United Launch Alliance, and several regional commercial networks and private SSA service providers.

18 SDS operates command and control for the U.S. Space Surveillance Network (SSN), a global collection of ground and space-based sensors that track all man-made objects in Earth’s orbit, from Vandenberg Space Force Base in California. 18 SDS is the primary global space object catalog authority. As such it maintains the resident space object (RSO) database of >46,000 natural or artificial tracked objects in orbit and runs the SSA sharing program that feeds data to US agencies, allies, and commercial satellite operators.

The worldwide Space Surveillance Network collects and processes real time data from a worldwide sensor network including phased-array radars, ground-based electro-optical sensors, deep‑space telescopes, and space-based sensors to detect launches, track objects, characterize breakups, and issue conjunction (collision) warnings to satellite operators to plan and execute collision-avoidance maneuvers.

Other National/Governmental SSA organizations:

• European Space Agency (ESA) – ESA, headquartered in Paris, France, co-ordinates the European Space Tracking (ESTRACK) network consisting of a number of ground-based space-tracking stations and facilitates communications between ground operators, and similar networks are run by the USA, China, Russia, Japan, and India.

• EU Space Surveillance and Tracking (EU SST) program - provides tracking, re‑entry prediction, and collision‑avoidance services for European space operators.

Commercial Ventures/Industrial Organizations:

There is a large eco-system of private and commercial organizations that provide space awareness, object tracking, and related services including Aerospace Corporation, BAE Systems, Beyond Gravity, ExoAnalytic Solutions, LeoLabs, NorthStar Earth & Space, (US Defense Primes: Lockheed Martin, Northrop Grumman, L3Harris, Raytheon), and European organizations Aldoria and Telespazio, just to name a few.

Net, net, SSA is a layered ecosystem where 18 SDS acts as the de facto global reference catalog provider, while other service providers deliver value added services, including monitoring, detection, tracking, and data analytics for operators and defense customers.

How do SSA tracking and automated collision avoidance systems work?

Satellite (and space object) tracking leverages combinations of both ground-base and space-based networks of radar and optical systems. SSA architectures deliberately combine radar’s ranging accuracy and all weather availability with optical senor’s high sensitivity and deep‑space reach.

The US’ Space Surveillance Network leverages radar systems for much of its ground-based sensor capabilities. Radar is typically used in detecting Low Earth Orbit (LEO) targets and it is robust technology that is largely unaffected by lighting, atmospheric turbulence, and weather conditions. Radar can accurately determine distance to a target and can track multiple objects simultaneously 24x7. A core issue with radar is its significant power requirements.

Quick reminder: Radar is an active sensor system that transmits radio waves and measures the reflected signal from an object to determine the object’s range, speed and trajectory which provides very accurate positioning.

Radar performance in SSA is strongest for low earth orbits (200-2000 km above sea level) and medium earth orbits, MEO, (2,000-35,000 km above sea level) where its signal strength is strong enough to be effective. Because radar signal strength falls off rapidly by the fourth power of distance (1/d to the 4^th) tracking objects above MEO with ground radar becomes prohibitively power and cost intensive.

Optical systems in SSA are passive sensors - they collect sunlight reflected or emitted by satellites and debris and are typically deployed using telescopes coupled with imaging detectors. This makes them comparatively cheaper and more scalable than radar arrays for covering high‑altitude objects like satellites in geostationary earth orbit (GEO). Ground‑based optical sensors are negatively affected by weather, atmospheric turbulence, and light pollution, so are most effective with clear, dark skies which limits their active duty cycle and field of coverage.

Space‑based optical (SB-SSA) overcomes many of the limitations of ground based systems by avoiding clouds and much of the atmosphere and therefore have longer illumination windows, but at the cost of more complex platforms and orbital design. And they are yet another satellite!

Beyond radar and optical, other sensor solutions including laser and infrared sensors, are typically less developed, more expensive, and/or more unique in their use-cases for SSA than their optical and radar counterparts.

In practice, SSA architectures use radar as the backbone for continuous LEO surveillance and event detection, then augment with optical systems - ground or space‑based - for GEO monitoring, detailed characterization, and follow‑up observations of high‑interest objects. This complementary design underpins many national and commercial SSA networks, reflecting the reality that “radar versus optical” is less a competition and more a deliberate division of labor across orbital regimes and mission needs.

It will come as no surprise to note that it is increasingly common for SSA observation and tracking systems to combine/merge/fuse both optical and radar data sets into a single analysis stream. While it sounds cool to have the ability to merge data streams, the sheer volume of these data streams, and the differing data characteristics create significant challenges for interpretation and analysis. Enter the application of AI and ML technologies to enable multiple data sources to be assessed in real time with greater depth and detail to derive the unique insights in support of SSA decision making.

Stepping Back - The Big Picture

Currently, there are no internationally agreed to formal rules of the road for Satellite Traffic Management (STM) or “deconfliction” coordination. Existing arrangements are ad-hoc, fragmented, and voluntary with no binding global traffic management authority. There are guidelines at the UN level, plus a patchwork of bilateral and regional space situational awareness data‑sharing agreements and industry best‑practice protocols.

The United States, through U.S. Space Command (USSPACECOM), has entered more than 185 Space Situational Awareness (SSA) information‑sharing agreements, with foreign governments, international organizations, and commercial satellite operators, which support collision avoidance and deconfliction but remain essentially bilateral or multilateral technical arrangements rather than a formal global traffic regime. Can you say – totally unmanageable!

How is it that private satellite operators are free to launch thousands of objects, literally thousands of floating objects, into space without a systematic method, process, or protocol for coordinating, communicating, and managing with other operators? Is it fair to suggest that private operators should bear the some of the costs/burden for operating a globally capable STM system… aka a dedicated air traffic control system for space? All good questions.

A robust global STM framework is critical for national and global security. An agreed and preferably enforceable framework can establish the technical, legal (regulatory) and operational behaviors required for safe and sustainable satellite operations. With the rapid growth of commercial satellites and satellite operators, the sheer volume of space traffic to monitor and track is overwhelming, let alone the ability to identify risks and risky behaviors.

It’s a matter of when, not If, we will have a collision of satellites in space with the potential for bits and pieces of space junk to rain down on our heads. We need a robust STM now.

Please share your thoughts via the Comments section for this post or open a new SubStack chat thread … and please forward this post to your friends and colleagues. See you next time in 2026!

Why this a hot space! The Quantum Internet!

It is envisioned that by faithfully relaying quantum states over large distances, repeaters could someday connect distant quantum processors/computers into a distributed network analogous to today’s global internet. In this way, quantum repeaters can enable the creation of a quantum-secure global communications backbone, a “quantum internet”, that could connect data centers, financial institutions, government facilities, with unbreakable encryption and quantum level security ensuring that adversaries cannot eavesdrop or interfere with critical communications. Quantum repeaters offer the promise of extending the advantages of quantum security to the scale of real global networks.

It is estimated that there are more than 150 projects worldwide working on quantum networking technologies and systems. This estimate includes national or regional quantum-communication programs (example: U.S. DOE quantum internet effort), academic and national-lab testbeds and pilots, campus/metro fiber loops, satellite demos, quantum-repeater test networks, and commercial/startups focused on quantum networking hardware, software stacks, and quantum communications. infrastructure.

​

What is a Quantum Repeater?

Quantum repeaters are specialized devices designed to address the fundamental challenge that qubits cannot be directly measured, copied, or inspected without leading to decoherence, aka collapsing their quantum state – per the no-cloning theorem.

Quantum repeaters leverage entanglement and quantum memory instead of measurement and amplification. Quantum repeaters break long communication distances into shorter segments and create entangled quantum states across each segment. Through a process called entanglement swapping (more on that below), these segments can be joined together extending the distance that qubits can be “transported”.

By using entanglement and quantum memories, repeaters relay quantum information across the network without directly transmitting photons end-to-end. Stated another way, repeaters “teleport” quantum states from one network segment to the next. Quantum memories at the repeater nodes perform a critical task by temporarily storing qubits, ensuring that entanglement can be established and synchronized across segments. Net-net, quantum repeaters enable long-distance quantum networking by overcoming photon loss and preventing decoherence, all while obeying quantum laws.

How a Repeater works - let’s go under the hood and detail the internal operations of quantum repeaters

As you might imagine, the operation of quantum repeaters is complex and involves multiple systems and sub-systems including:

• Entangled photon sources – to generate entangled photon pairs generation at rates from kilohertz to megahertz.

• Entanglement Swapping – used to extend entanglement across multiple segments. (more below)

• Quantum Memories – with control lasers and magnetic or optical fields to store, rephase, and retrieve the photon’s state on demand to enable synchronization.

• Error Correction – or entanglement purification – techniques are applied to improve the quality of the entangled states by reducing noise and errors.

• Light–Matter Interfaces – where one photon of each pair is mapped into a memory while its partner travels down the fiber.

• Quantum Frequency Converters and Filters – required to adapt wavelengths and clean up spectral modes

• Control Electronics and Software – lots and lots of advanced (often proprietary) software to coordinate signals, timing, and operations.

These components and systems are synchronized so that when detectors signal (“herald”) that entanglement was successfully created on a link, the corresponding qubit is kept in memory while neighboring links repeat the process until the entire chain is ready for entanglement swapping. The process of entanglement swapping, purification, and storage is repeated along the chain of nodes until entanglement spans the entire network distance between the source and destination.

Photon Sources

Photon sources generate single photons or entangled photon pairs to enable entanglement swapping that extends quantum links over long distances. Photon sources emit light particles with precise quantum properties, like narrow spectral lines, to interact efficiently with quantum memories in repeater nodes. They often produce pairs where one photon suits the memory system being used in the repeater and the other travels in telecom fibers.

Common Photon Source Types include:

• Quantum Dots: These are semiconductor nanostructures that emit highly indistinguishable single photons or pairs, often enhanced by cavities for better extraction. They operate in near-infrared or telecom bands ideal for networks.

• Nitrogen-Vacancy (NV) Centers in Diamond: NV defects in diamond cavities act as sources by mediating entanglement between photons and spins, supporting telecom wavelengths via frequency converters.

• Erbium-Based Emitters: Doped materials that naturally emit photons at telecom wavelengths (around 1532 nm), matching fiber transmission without conversion.

Entanglement Swapping

Entanglement swapping is a foundational capability in repeaters to extend entanglement across multiple network segments. It is a process of taking two independent entangled photon pairs and performing a joint measurement (a Bell State Measurement) that “swaps” the entanglement, resulting in two previously unrelated particles becoming entangled.

A Bell State Measurement identifies exactly which of the four standard entangled patterns (Bell states) they share. This identification enables quantum communication and computing protocols to leverage entanglement. The outcome of the Bell state measurement can be communicated to the end nodes using standard network methods, so they know their particles are entangled, this is referred to as heralding. This is how a repeater chains multiple network segments together, by entangling neighboring nodes, then migrating the entanglement through the nodes until the end nodes share a state. This migration of quantum states from one place to another without moving the physical particle itself is referred to as “quantum teleportation”.

Quantum Memories

Each repeater node needs reliable quantum memory. Quantum memory is a critical component because entanglement distribution is probabilistic, aka it’s a somewhat random process, and is sometimes slow. Given there are typically multiple segments in a network, a repeater needs to be able to hold on to the entanglement in one segment while waiting for the others to be established, so then entanglement swapping can be performed. Quantum memories perform the role of synchronizing the entanglement distribution process, making it feasible to string together many short network segments into one long entangled network connection.

Error Correction or Entanglement Purification

As described in our foundations piece on Quantum Technologies quantum information is notoriously fragile and subject to errors and noise. There are several error correction methodologies and techniques used in repeaters. See this post for more details on error correction.

Entanglement purification (also known as distillation) is a correction protocol where two nodes share multiple low-fidelity (aka noisy) entangled pairs and then apply measurements and classical communication to “distill” a single pair with higher fidelity (aka less noisy). The trade-off here is purification requires two-way communication between nodes and multiple entangled pairs, which can significantly slow down processing time and negatively affect throughput rate of the repeater.

Additional error correction approaches actively being researched include protocols for all-photonic repeaters that correct losses on the fly using quantum error correction codes. These systems will likely need to implement small quantum processors at each node to actively correct errors – aka they are more complex and expensive to build.

Uncorrected, each segment in a long network chain will degrade entanglement quality, and after many hops the entanglement could become too noisy to be useful. The use of error correction techniques is therefore essential for scaling long distances and it is a core capability in quantum repeaters.

Projects/TestBeds/Programs

Quantum repeater technology is rapidly progressing in research labs, and test beds, but reliable widespread practical deployments have not been achieved. Academic institutions, government sponsored programs and private companies are highly active in working to develop a global quantum network. Here are just a few examples:

Academic:

• QuTech at Delft University in the Netherlands demonstrated (in 2021) a multi-node quantum network, (nodes were 1.3km apart), entangling three remote quantum processors. This lab based demonstration showed that a small quantum network with a repeater-like node is feasible.

• German researchers demonstrated the distribution of entanglement between a single-atom-based quantum network node and a photon transmitted over up to 101 km of telecom fiber.

• Researchers from Harvard and MIT, in 2024 in the Boston area, built a two‑node quantum network using silicon‑vacancy centers in nanophotonic diamond cavities over up to 40 km of spooled telecom fiber and a 35 km deployed urban fiber loop.

• In Europe, in 2023, researchers at the University of Innsbruck (Austria) and France’s Paris-Saclay unveiled a prototype quantum repeater system that transmitted entanglement over 50km using an intermediate repeater node​.

Government Sponsored Project/Programs

• The European Union’s Quantum Internet Alliance includes groups from Delft, Paris, Geneva, and others, working toward a blueprint for a pan-European quantum network. There are multiple testbeds in the EU connecting cities.

• In early 2025 the Canadian Federal Government announced more than CA$74M to support 107 quantum projects across computing, communications, encryption, materials and sensing, funded through Alliance Quantum, Consortia Quantum, and International Quantum grants.

• Japan funds quantum technologies through several large national programs, each containing multiple individual projects in quantum computing, communications, sensing, and basic science. Major projects include MEXT Q-LEAP (Quantum Leap Flagship Program), Research and Development of Superconducting Quantum Computers (RIKEN), JST “Moonshot” R&D Program, and JST Strategic Basic Research programs. ​

• The USA has too many active programs and projects in the field of quantum networking to list here – see this summary list for a current snapshot.

Private Companies – there is a ton of activity in the commercial space in the USA. Here are a few notable players.

• Startups and venture backed companies – notable commercial startup companies include D-Wave, IonQ, Photonic, PsiQuantum, Rigetti , Riverlane, and QuEra, just to name a few.

• Amazon established the AWS Center for Quantum Networking (CQN), and in 2023 partnered with Harvard University to demonstrate a quantum network over 35km of deployed fiber​. CQN complements the advanced quantum science and engineering efforts already underway at the AWS Center for Quantum Computing and the Amazon Quantum Solutions Lab.

• Networking stalwart Cisco has a quantum networking research group working on repeater protocols, software, and hardware integration​. In November 2025, IBM and Cisco announced a collaboration to establish a network of large-scale, fault-tolerant quantum computers, combining IBM’s quantum computing with Cisco’s quantum networking solutions targeted for realization in the early 2030s.

• Google’s quantum efforts center on building fault-tolerant superconducting quantum computers, powered by Google semiconductors (Willow) while also developing software tools and collaborations around those devices.

• Microsoft is offering Azure Quantum cloud platform, a managed Azure service that lets users run quantum programs on real hardware from partners such as IonQ, Quantinuum, PASQAL, and Atom Computing. In 2025 it introduced the Majorana 1 chip, integrating multiple topological qubits on a tile layout intended to scale to around a million qubits and support trillions of operations for chemistry and materials problems

• Honeywell joined existing investors in Quantinuum (trapped-ion technology) with a $600M equity capital investment to advance cloud based quantum hardware-as-a-service offering, and an on-premise offering called Helios.

• Not surprising is NVIDIA’s has multiple efforts in quantum and are focused on hardware (GPUs), software, and interconnects that to help make quantum devices usable and scalable. CUDA‑Q is Nvidia’s quantum programming platform that lets developers write hybrid CPU–GPU–QPU workflows. The cuQuantum SDK provides tensor‑network and state‑vector libraries optimized for Nvidia GPUs, enabling high‑fidelity simulation of tens to hundreds of qubits. Nvidia also offers NVQLink a low‑latency, high‑bandwidth interconnect system that attaches QPUs and quantum control systems directly to Nvidia GPU servers.

Remaining Challenges

Quantum networking and its enabling technologies, repeaters, error correction are a hot bed of activity. Naturally, as with all things quantum, there are complex challenges still to be overview. Here are a few challenges.

• Loss and Distance: Optical fiber has readily identifiable loss characteristics and while repeaters can address these losses, repeaters also introduce loss and error probabilities. This requires that repeaters be placed at optimal intervals and the number of repeaters required can grow quickly.

• Quantum Repeaters Complexity: Repeaters are complex quantum devices, far more complicated (and expensive) than a classical amplifier. Designing and implementing this infrastructure is a multi-dimensional engineering effort.

• Error Rates/Reliability: Current quantum hardware is not as reliable as classical networking equipment. To scale up, quantum repeaters must become orders of magnitude more reliable in their operation. This challenge is similar to that faced with building quantum computers - error rates must be lowered or correction systems need to be more robust.

• Cost/Economies of Scale: There remains a massive amount of engineering needed to develop systems and sub-systems that can be manufactured at scale. Today’s quantum repeater deployments can best be described as bespoke custom implementations and test bed experiments.

• Integration with Existing/traditional Network Infrastructure: Integrating quantum channels with existing fiber networks where quantum signals will co-exist in the same fiber bundles as traditional signals creates unique challenges primarily related to cross-talk and interference. Further, given the fragile nature of quantum information, ensuring quantum hardware can operate reliably in the real world is a major hurdle.

Wrapping Up

In the world of quantum repeaters, the current status is that quantum repeaters are transitioning from theory to practice. We have isolated demonstrations of all the necessary pieces: long-lived quantum memories, entanglement swapping operations, entanglement over long distances, and even small quantum networks. The challenge now is engineering these components into integrated systems that can operate outside a lab. Organizations around the world are heavily invested in this – from university labs pushing the science, to startups building hardware, to government-funded networks acting as testbeds.

The combination of entanglement swapping, robust quantum memories, and error correction/purification protocols will ultimately allow quantum repeaters to deliver high-quality entanglement over arbitrarily long distances. These are the building blocks of tomorrow’s quantum internet infrastructure.

Thanks for reading. Let us know what you think. Please share your thoughts via the Comments section for this post or open a new SubStack chat thread … and please forward this post to your friends and colleagues. See you next time!

By temporarily storing quantum states, quantum memory serves a critical role in teleportation-based quantum repeaters which enable long distance quantum communications beyond the loss limitation in optical fibers. More on Quantum Repeaters in a future post! Historically, quantum memories have been developed along two distinct paths: optically controlled memories and engineered absorption.

Solid State and Optical Systems

Solid state systems. Solid state quantum memories store quantum information in solid-state systems, which are typically easier to integrate into electronic devices compared to other forms of quantum memories, such as those based on atomic systems. This type of quantum memory includes nitrogen-vacancy (NV) centers in diamonds. NV centers leverage relatively easy optical preparation systems and readout (retrieval) of electronic and nuclear spin states along with relatively long coherence times – an important consideration for quantum memory.

Optical systems. Optical systems for quantum memories involve using light to store and manipulate quantum information. An example of this type of memory is rare-earth ion doped crystals that use quantum properties of light to store and retrieve quantum information. Rare-earth ion-doped crystals exhibit excellent coherence properties at cryogenic temperatures. This is particularly important for application of quantum memories in long-distance quantum communications that require storage times greater than milliseconds.

Rare-earth ion-doped crystals have initially been used for implementing storage of time-bin qubits using the AFC protocol. AFC (Atomic Frequency Comb – more below) is a quantum memory technique that stores photonic qubits in a non-homogenous solid-state media with comb-like structure. The stored quantum data is re-emitted after a fixed, programmable time set by the comb spacing.

Absorption Systems

Atomic ensembles use a large group of atoms as a medium to store quantum information. This approach leverages the collective quantum states of atoms to create a robust and efficient storage mechanism for quantum information. These systems are distinct from solid-state quantum memories in that they typically use gases or vapors of atoms as the medium for quantum storage. Examples of atomic ensemble quantum memories include rubidium vapor cells, cavity QED (quantum electrodynamics), and cold atoms using electromagnetically induced transparency (EIT).

• Quantum Memory using Rubidium vapor cells

These systems store and retrieve quantum states of light, typically using rubidium gas atoms in warm or room-temperature that exist in superpositions of multiple energy states simultaneously, which can be influenced by electromagnetic fields. Compared with cold-atom memories, warm rubidium cells are simpler to assemble and operate at or near room temperature, which is advantageous for deployment in real-world quantum networks and field tests. They are also compatibility with existing fiber networks, making them a leading approach for quantum repeater nodes to enable long-distance and quantum communication links.

• Cavity Quantum Electrodynamics – Cavity QED

Cavity QED approaches are actively being developed for quantum memory systems because they enable strong and stable (coherent) interactions between light and matter. Cavity QED uses resonant optical or microwave cavities to confine electromagnetic modes and enhance interactions with quantum emitters such as atoms, ions, or artificial qubits.

An optical cavity, also called a resonating cavity or an optical resonator is an arrangement of mirrors or other optical elements that traps light at specific resonant frequencies so that photons bounce back and forth many times and strongly interact with a quantum emitter (such as an atom, ion, quantum dot, or superconducting qubit). By matching the cavity resonance to an internal transition of the emitter, the electromagnetic field is confined to a small mode volume which boosts the strength and coherence of light–matter interactions beyond what is possible in free space.

• Circuit Quantum Electrodynamics - Circuit QED

Another approach known as Circuit QED is also exploring interactions between light and matter where a single photon within a single mode cavity coherently couples them to a quantum atom. In Circuit QED the photon is stored in a one-dimensional on-chip microwave resonator and the quantum object is not a natural atom but an artificial one.

• Electromagnetically Induced Transparency (EIT)

EIT is an optical phenomenon in atoms that uses quantum interference to induce transparency into an otherwise resonant and opaque medium. EIT systems create and retrieve the stored excitation without destroying the quantum state, and can be used for slow light, light storage and quantum memory applications.

EIT is a quantum interference phenomenon in three-level systems that generates a narrow transparency window, enabling impactful control over light propagation. It utilizes a strong control field alongside a weak probe field to create destructive interference that minimizes absorption and enhances nonlinear optical effects.

EIT-based systems are preferred for long storage times, ease of implementation, and situations where high-fidelity operation is essential without demanding bandwidth requirements. They are especially useful when low control field power and simple setup are priorities. EIT underpins diverse applications including quantum memories, switchable photonic devices, high-resolution spectroscopy, and slow-light propagation in various platforms.

• Raman-based systems

These systems excel in high-bandwidth, high-speed quantum memories, accommodating ultrashort pulses and higher repetition rates. They have recently demonstrated near-unity efficiency and fidelity, rivaling the best EIT systems, although they require careful control of technical noise and strong drive fields.

Overall, the choice between EIT and Raman depends on the needs: for broadband or high-speed storage, Raman-based cavity schemes are emerging as the performance leaders; for established high-fidelity but narrower bandwidth quantum memories, EIT remains a robust standard.

Atomic Frequency Comb (AFC) Storage Protocol

AFC memory is a quantum memory technique that stores photonic qubits in inhomogeneously broadened ensembles by preparing an absorption profile with a comb-like structure.

Inhomogeneously broadened ensembles refer to large groups of atoms or ions whose optical transitions have a range of frequencies, rather than a single well-defined frequency. This enables AFC systems to store quantum states of light by collectively absorbing photons as coherent excitations spread across many atoms with different resonance frequencies.

When a photon is absorbed, it creates a collective excitation across the atomic ensemble, and re-emits after a fixed, programmable time set by the comb spacing. It is particularly well-suited for multimode and telecom-band storage in solid-state media.

Wrapping up

This article simply scratches the surface of topics related to quantum memory systems and technologies. Quantum memory is a foundational, and EXTREMELY COMPLEX set of technologies required for implementing advanced quantum networks, quantum computing synchronization, and practical long-distance quantum communication.

Researchers around the globe are working to maintain quantum coherence in quantum memory over time by addressing environmental noise and decoherence, balancing storage time (coherence) with bandwidth and retrieval speed, and integrating memory systems with other quantum systems, as well as improving the efficiency of photon-memory coupling, and improving scalability.

Thanks for reading. Let us know what you think. Please share your thoughts via the Comments section for this post or open a new SubStack chat thread … and please forward this post to your friends and colleagues. See you next time!

Error Correction Founding Principles

Simply stated, QEC combines multiple physical qubits with software to form logical qubits that are more robust, allowing errors to be detected and corrected without disrupting the encoded quantum information. It is possible to spread the (logical) information of one logical qubit onto a highly entangled state of several (physical) qubits.

Peter Shor first discovered this method of formulating a quantum error correcting code by storing the information of one qubit onto a highly entangled state of nine qubits in 1995. Shor’s code demonstrated how to encode a single quantum bit across multiple physical qubits so that bit-flip and phase-flip errors could be corrected without collapsing the quantum state.

Physical Qubits

A physical qubit is a basic quantum bit implemented in hardware. A physical qubit can be realized by manipulating and measuring quantum particles or states that can exist simultaneously in a superposition, enabling quantum computation. Physical qubits can be created using a variety of techniques including trapped ions, neutral atoms, superconducting circuits, or single photons.

Logical Qubits

A logical qubit is a higher-level abstract qubit encoded across many physical qubits to protect the quantum information against noise and errors. Unlike a physical qubit, which represents a single quantum hardware entity, a logical qubit leverages several physical qubits entangled together so that errors in individual physical qubits can be detected and corrected, making logical qubits essential for fault-tolerant, scalable quantum computers even in the presence of noise and decoherence.

Implementing logical qubits requires significant overhead in terms both hardware and algorithm design given the number of physical qubits required to support logical qubits. The precise ratio of physical to logical qubits varies depending on the quantum error correction code and technology, but, it can often be on the order of hundreds or even thousands of physical qubits per logical qubit.

Overview of typical errors – bit-flip and phase-flip

QEC systems implement three types of error correction: bit-flip correction, phase-flip correction, and bit-phase flip correction.

• Bit-flip error correction takes on errors where a qubit’s state changes from 0 to 1 or the other way around. This type of error is like flipping a binary bit in classical computing. QEC codes detect and fix this error by encoding a logical qubit across multiple physical qubits and measuring parity checks.

• Phase-flip error correction deals with errors in the relative phase of a qubit’s state, where the phase in a superposition state changes. This kind of error is unique to quantum computing systems and can’t be directly measured without collapsing the state. To detect and correct phase-flip errors, the phase information is encoded redundantly across multiple physical qubits.

• Bit-phase flip error correction combines these two types by flipping the qubit’s value and inverting its phase at the same time. This combined error is corrected by integrating the methods of both bit-flip and phase-flip corrections.

An important consideration for QEC is the Threshold Theorem, which postulates that errors can be corrected effectively if the physical error rate remains below a certain threshold. By applying quantum error correction schemes, we can suppress the logical error rate to arbitrarily low levels. Stated in plain English - the Threshold Theorem indicates you need to be able to correct errors faster than they are created – otherwise the computation system will degrade and be destroyed by noise.

While estimates vary widely and are difficult to calculate due to the difficulty of simulating large quantum systems, current estimates put the error threshold for surface code implementations in the order of 1,000 to 10,000 physical qubits per logical qubit, if physical qubit error rates stay below roughly 1%. This 1% error rate directly impacts the volume of required physical qubits to be effective, and research continues to try and find ways to reduce both the error thresholds and the qubit overhead. Today, practical full-scale applications may require millions of physical qubits for robust quantum error correction in large-scale devices.

How it works – Overview of various QEC schemes/mechanisms

Researchers and commercial enterprises are exploring several QEC schemes that encode one logical qubit of quantum information into a larger entangled state of many physical qubits in such a way that certain errors can be detected and fixed. There are various families of QEC codes, each with pros and cons. These systems all have these basic mechanisms.

• Redundancy – multiple physical qubits per logical qubit,

• Syndrome (error) measurements – extra operations that periodically check for signs of errors without directly measuring the logical information

• Decoding and correction – a method, often algorithmic or hardware-based, to interpret the syndrome signals and decide how to correct the system if an error is detected.

Here are some of the main techniques being worked on with each method implementing logical qubits but in different physical ways, with different trade-offs in overhead and performance.

Surface Codes

This is a leading QEC technique which encodes logical qubits using physical qubits arranged in a 2D lattice. and uses local interactions between neighboring qubits to check for errors. In a surface code, a logical qubit is encoded in the joint state of a dxd patch of physical qubits (d is called the code distance). Surface codes correct errors by measuring correlations (syndromes) without directly measuring qubit states. Each surface code represents a single encoded or “logical” qubit.

For example, a distance-3 surface code might use a grid of data qubits (plus some ancillaries for measurement), and a distance-5 code uses a grid. Surface codes have the benefit of a relatively high error threshold – around 1% error per gate in the physical qubits and increasing the code size will exponentially suppress logical errors. The bigger a surface code lattice, the more errors it can tolerate and the logical qubit is more and more protected, and the logical performance improves.

Surface Code Distance 7 Lattice: Source: https://www.nature.com/articles/s41586-024-08449-y/figures/1 https://www.nature.com/articles/s41586-024-08449-y

Color Codes

Similar to surface codes, color codes, another class of 2D topological codes also arrange qubits on a lattice, often using triangular tilings, and has check operators that involve multiple (3 or 4) qubits. The physical qubits are placed on the vertices of a lattice that is made of triangular tiles. These lattices are trivalent graphs, meaning each vertex connects exactly three edges, and the faces (the triangles) can be colored with three different colors (e.g., red, green, blue) so that no two adjacent faces share the same color. This three-colorable tiling is important because it allows the definition of stabilizer operators used to detect and correct errors on each colored face.

Low-Density Parity-Check (LDPC) Codes

Recently, highly efficient and scalable quantum LDPC codes have been developed that can handle hundreds of thousands of logical qubits and they simultaneously correct both bit-flip and phase-flip errors.

The “low-density” aspect means that each parity check only involves a small number of qubits, and each qubit only participates in a few such checks. While surface codes are more hardware-friendly but qubit-expensive, LDPC codes can obtain higher encoding efficiency and lower overhead at the cost of more complex qubit arrangements.

LDPC codes can be viewed as an advanced class of QEC that includes surface codes as a special case. LDPC go beyond surface codes to improve scalability and resource use efficiency, however LDPC codes are a less mature technology, with active research exploring practical implementations and decoding strategies.

Calderbank-Shor-Steane (CSS) Codes

These codes are a class derived from classical error-correcting codes and are widely used in quantum error correction schemes. CSS codes combine two classical error correction codes to detect and correct two main types of errors: bit-flip errors (where a 0 becomes a 1 or vice versa) and phase-flip errors (a quantum-specific error related to the phase of the qubit).

They are built from two classical codes that should satisfy the “dual-containing” condition, meaning one code’s dual is contained inside the other. Because CSS codes use classical decoding algorithms, they make the process of correcting errors in quantum computers more practical and scalable, especially with large codes.

4D Geometric Codes

Microsoft researchers are actively developing 4D quantum error-correcting codes, which embeds quantum information in a lattice structure with four spatial dimensions, aka 4D, rather than the usual two dimensions used in standard surface codes. This higher-dimensional arrangement creates many more pathways to detect and fix errors efficiently, reducing errors by up to 1,000 times compared to traditional 2D codes.

The 4D structure enables “single-shot” error correction, meaning errors can be identified and corrected in a single round of measurements, speeding up computations and simplifying how the quantum computer operates. Unlike typical methods that require multiple cycles of checking, these codes catch errors quickly without repeatedly interrupting the calculation.

In simple terms, 4D quantum error-correcting codes need far fewer physical qubits to represent each logical qubit, and wrap qubits in a complex, four-dimensional shape that helps catch errors quickly and accurately, allowing quantum computers to compute more reliably and efficiently with less delay.

Cat Codes

Named after Schrödinger’s cat (the famous thought experiment), cat codes encode a logical qubit into superpositions of coherent states in a microwave cavity – essentially “alive” equals a 0 state and “dead” equals a 1 state of superpositions of an electromagnetic field.

In cat codes, these states are designed so that errors, particularly the loss of a single photon, change the parity of the photon number, a property that can be detected without disturbing the encoded quantum information. This allows for correction of photon loss errors by measuring the photon number parity using an ancillary system coupled to the quantum state.

GKP Codes

The GKP code encodes a qubit into continuous variables of a quantum harmonic oscillator, such as the position and momentum of a mode of light or a cavity. Imagine the quantum state as a point on a graph with a grid pattern. Normal small mistakes in the system just nudge the point slightly off the grid. The GKP code detects how far off the grid the point is and a corrective shift is applied to bring the state back exactly onto the nearest grid point, thus correcting the error, and protecting the quantum information stored there from errors.

Bosonic Codes

This advanced method encodes logical qubits into a single bosonic mode, aka a quantum harmonic oscillator with infinitely many energy levels, rather than spreading information across many physical qubits.

A quantum harmonic oscillator describes a particle that moves back and forth in a way similar to a mass on a spring, but with important quantum properties where the energy has specific, discrete energy levels that the particle can occupy.

Instead of duplicating qubits many times like in traditional QEC methods, bosonic codes use a mode (e.g. an electromagnetic mode in a microwave cavity) that can exist in many photon number states to represent and protect quantum information. These codes detect common errors like photon loss by measuring patterns such as photon number parity, which reveals errors without revealing the encoded information itself.

Advantages of bosonic codes include requiring fewer physical components and the ability to handle the dominant error type (photon loss) making error correction hardware efficient. Disadvantages include the need for high-quality oscillators (very long-lived cavities or modes), and implementation challenges of applying multiple gates between logical bosonic qubits and other factors that are computationally intensive. Examples of bosonic codes include the Cat Code, GKP Code, and Binomial Code.

Challenges

Scaling Physical Qubit Count An obvious challenge is the sheer number of physical qubits required. Building a chip with a thousand qubits is one thing; getting all thousand to perform well simultaneously is another. Today’s largest quantum chips have on the order of a few hundred qubits. To run even a modest algorithm with full error correction, we likely need thousands or millions of qubits. The jump from 100 qubits to 1,000,000 qubits will require innovations at every level, from nanofabrication and wiring to software and error decoding.

Improving Physical Qubit Quality Although error correction can reduce error rates, having better starting material (physical qubits) makes everything easier. There is a virtuous cycle here: the lower the physical error per gate, the smaller the code can be to achieve a given logical reliability. Many researchers are simultaneously trying to improve qubit quality through advanced materials research, better fabrication, and novel qubit designs like fluxonium or dopant spins.

Real-Time Low Latency Decoding A key component to QEC is decoding. During QEC, measurement data, which is called a syndrome is acquired, which gives clues as to where errors happened during computation. Now that we have captured and identified error syndromes it is up to a decoder to quickly interpret the measurements, identify the most likely error, and determine the corresponding correction. Sound logical and simple – right!

The latency of the decoding process, known as the “decoding response time”, sets the logical clock rate for the execution of fault-tolerant quantum algorithms. If decoding is slow, error data can accumulate, creating a backlog that slows down the speed of computation of the entire system. Experiments have demonstrated that to be effective the entire decoding system must process error syndromes and conditionally execute correction gates within microseconds.

The throughput of a real-time decoder—the rate at which it processes syndrome data—must always exceed the rate at which data are generated by the quantum processor to prevent bottlenecks. Parallelization of decoding and low-latency hardware integration are approaches used to avoid computation slowdowns.

We need decoders that can handle maybe 1000 syndrome bits every microsecond, and output corrections in real-time. Researchers are exploring FPGA and ASIC implementations of decoders, as well as improved algorithms, including machine learning approaches to decoding. The challenge is to keep the decoding fast enough that it doesn’t become a bottleneck.

Cross-Technology Integration To scale quantum systems it will likely require hybrid approaches that integrate physical qubit systems with bosonic cavities for memory, and photonic links for communication. Ensuring the whole system remains fault-tolerant across these interconnected systems is non-trivial. We can expect modular approaches with groupings (modules) of component technologies integrated into large scale systems.

Summary

Certainly, there are many challenges that need to be addressed to create a fully functioning, stable, fault tolerant quantum network. Scalable low latency error correction is one. Reliable quantum memories is another. Another important missing piece is the ability to extend the reach of quantum links to arbitrarily long distances, using quantum repeaters. Quantum states cannot be simply copied and regurgitated, as is done with classical information. Quantum nodes will need sophisticated quantum logic gates to ensure that entanglement is preserved in the face of losses from interaction with the environment.

Thanks for reading. Let us know what you think. Please share your thoughts via the Comments section for this post or open a new SubStack chat thread … and please forward this post to your friends and colleagues. See you next time!

Quantum networking leverages the unique properties in quantum mechanics of superposition and entanglement to transmit quantum information (qubits) between devices. Quantum networks hold the potential to enable fundamentally new capabilities, including ultra-secure communication with unbreakable encryption, distributed quantum computing, and hold out the promise of creating a “Quantum Internet”.

Secure communications

Today’s communications networks are vulnerable and subject to hacking and intrusion. Quantum networks can enable completely new forms of security that relies only on the laws of physics and is unbreakable and completely hack proof. Quantum networks establish connections using entanglement, rather than passing data directly through the network. In addition, because of the properties of quantum states, it is possible to detect with certainty if communication has been intercepted.

How it works

The basic structure of a quantum network and more generally a quantum internet is analogous to a classical network. In very simple terms, there are end points, communication links/infrastructure with repeaters to overcome limitations due to distance, and degraded performance caused by noise and other degradations.

First, we have end nodes/end points on which applications are ultimately run. These end nodes are quantum processors of at least one qubit. Some applications of a quantum internet require quantum processors of several qubits as well as quantum memory at the end nodes.

Quantum networks leverage qubits and because they can exist in superpositions of 0 and 1, so they make quantum transmissions more complex than classical bits. Also, because qubits can be entangled, where the state of one instantly affects the other, even across long distances, quantum networks leverage entanglement for secure communication.

In quantum networks Qubits are typically encoded into photons. Photons enable transmission and can leverage existing fiber optic infrastructure – a HUGE positive! Entangled photon pairs are generated at specialized sources, where two photons become quantum mechanically correlated. Entangled qubits have no predetermined states until measured. When one photon’s quantum state is observed, it instantly determines the corresponding state of its entangled partner.

There are a variety of photon encoding technologies including:

• Polarization Encoding: The polarization state of a photon (such as horizontal vs. vertical or diagonal orientations) serves as the logical 0 and 1 for the qubit.

• Path (or Dual-Rail) Encoding: This scheme uses two distinct optical paths for a single photon. The photon’s presence in one path is a logical 0, and the other is a logical 1.

• Time-Bin Encoding: Qubit states are encoded based on the time a photon arrives at a detector. Typically, one time slot is 0, and another delayed slot encodes 1; superpositions involve the photon being in a quantum mixture of both times.

• Frequency Encoding: Different frequencies (colors) of photons are used to represent different qubit states, with superpositions involving photon states across different frequencies.

​Additional more complex encoding schemes use squeezed light, “cat” states or Gottesman-Kitaev-Preskill (GKP) states or amplitude and phase quadratures of electromagnetic fields. These schemes are typically used to protect quantum states in unique ways to develop resilience to certain errors.

​Deployment over fiber optic networks. To transport qubits from one node to another, we need communication links. For the purpose of quantum communication, standard telecom optical fibers can be used. Optical networks have the advantage of reduced chances of decoherence. Typically, photons (particles of light) serve as carriers of qubits, sent through optical fibers or free-space links (see below for more). As the photons travel through fiber, new specialized technologies are required to preserve qubit states and entanglement.

Quantum repeaters. In fiber-optic networks transmission loss is compensated for by introducing amplifiers, which boost weak signals by producing many copies of the input photons. However, due to the no cloning theorem, an optic amplifier cannot perfectly copy a quantum state. The noise produced in the process of amplification destroys the entanglement necessary for quantum communications.

To overcome distance limitations and keep entanglement, quantum repeaters are used. Repeaters appear in between end nodes. Since qubits cannot be copied (no-cloning theorem), classical signal amplification is not possible. By necessity, a quantum repeater operates in a fundamentally different way than a classical repeater. These devices extend quantum communication by relaying entangled states over multiple short hops, enabling connections beyond the range of direct transmission.

As noted above, quantum repeaters are needed to overcome distance limitations in a quantum network. Repeaters perform entanglement swapping to reliably extend the distance between which two devices can become entangled. Quantum repeaters correct for photon loss without disrupting the quantum state of the communicated information by catching and storing (rather than measuring) the quantum bits encoded in photons.

Quantum networking integrations into the fiber-optic network infrastructure is important for ensuring privacy in optical communications. Multi-core fibers (MCFs), the likely building blocks of future high-capacity optical networks, offer new opportunities for such integration.

Other deployment methods

Direct free-space distribution. Entangled photon pairs are sent directly between endpoints on the Earth’s surface or between satellites and ground stations. Ground based experiments have demonstrated successful transfer of entangled photons over several kilometers in open-air, even though turbulent and noisy atmospheric conditions.

Satellite based systems. By equipping satellites with entangled photon sources and high-precision pointing, it is possible to distribute entanglement across long distances (1,200 km) between ground stations. Entangled photons are generated on a satellite and transmitted directly to two separate ground stations, or vice versa. These systems implement narrow-beam divergence and advanced pointing and tracking techniques to maximize efficiency and overcome losses from diffraction and atmospheric absorption. Systems involving multiple satellites and ground stations are under development to further extend the range of these techniques. Satellite based systems are important in that they make it possible to establish secure quantum links globally, beyond fiber’s reach, and may well form the backbone of the futuristic “Quantum Internet”.

Hybrid satellite + fiber networks. Research is also making progress on systems that combine satellite-based distribution for long-haul connections with ground based fiber and repeater networks for regional links. Optimization algorithms are used to determine which nodes and links should actively relay or receive entangled photons.

Quantum key distribution (QKD)

Quantum networks can enable Quantum Key Distribution protocols, that provide encryption keys that are physically impossible for an eavesdropper to copy or intercept without detection. This is made possible due to the fundamental properties of quantum mechanics, including quantum superposition, entanglement, and where the process of measuring a quantum system disturbs the system which is easily detected. This unique property ensures that the distributed keys remain secure, as any attempt at interception will be immediately apparent and will invalidate the exchanged key.

This domain is seeing a lot of research and commercialization work. The development of QKD systems has reached a high level of technical maturity, with multiple commercial vendors producing products tailored for various applications. New protocols, integration with machine learning, and practical deployment over fiber and satellite networks characterize recent progress. Further, researchers are working on real-world implementation, multi-user systems, and addressing scalability, noise resistance, and security vulnerabilities.

Core QKD protocols. There are several different methodologies for quantum key distribution. The BB84 protocol is a seminal QKD method first introduced in 1984.

• BB84 is the most established and is a simple QKD protocol, relying on single photons with polarization encoding, and each photon representing a bit of data (zero or one). It achieves security by detecting eavesdropping via disturbances produced during measurement of non-orthogonal states.

• E91 uniquely leverages quantum entanglement and Bell inequality violations, giving it device-independent security not reliant on trust in devices, but it has more implementation complexity and lower key rates.

• B92 simplifies state choices to reduce implementation overhead, and continuous variable protocols encode information in quadrature amplitudes for compatibility with standard telecom technologies.

QKD protocols is a deep dark complex topic, and it deserves a dedicated post. More on this in the future. 😊

Research and testing implementations

Development in the quantum networking space is undergoing hyper acceleration on a global scale. There are way too many research, testing and commercialization efforts underway to list here. Suffice to say, quantum network test beds worldwide are transitioning from laboratory prototypes to integrated, field-ready systems supporting distributed quantum computing applications spanning university campuses, metropolitan areas, regional networks, and satellite links.

It is also not surprising that many of most advanced quantum networking research and commercialization activities combine university-led experiments, national test beds, and aggressive commercialization moves by industry players. Here are a few notable projects:

United States

Purdue University Quantum Network Testbed. Connects three laboratories via optical fiber underground, distributing photonic entanglement. Enables quantum key distribution (QKD), quantum process tomography, and microwave photonics over fiber.

Oak Ridge National Laboratory (ORNL) Quantum Communications and Networking Group. Maintains a metropolitan-scale test bed on a “dark fiber” network over 300+ km. Supports 400-gigabit conventional data transfer plus dedicated quantum channels.

The MIT Lincoln Laboratory Quantum Network Test Bed is a collaborative research initiative focused on building a scalable, high-rate quantum network for real-world testing of quantum networking applications. The test bed consists initially of two 43-kilometer optical fiber links connecting the Laboratory with MIT facilities.

Energy and Telecommunications Organization: EPB Quantum Network (Chattanooga, Tennessee). Commercial-scale quantum network where paying subscribers get access to an entanglement-based network with over 10 quantum nodes, equipment hubs, and operational control centers utilizing existing fiber infrastructure.

The U.S. Army Research Laboratory (ARL) and Air Force Research Laboratory (AFRL) have multi-node entanglement distribution and terrestrial/space-based test beds. These projects focus on both open research and applied tests with fiber and satellite links.

The U.S. Department of Energy (DOE) sponsors Quantum-in-Space collaborations, working with commercial vendors IonQ and Honeywell to leverage microgravity and advance quantum-secure communications, sensing, and scalable quantum networking.

Additional programs and test beds are operating at CalTech, Los Alamos National Labs, Argonne National Lab, Fermilab, and Brookhaven National Labs to name a few.

European Test beds

European test beds, around 15 have been reported across all major countries in Europe are often supported through the EU Quantum Flagship program. The European Commission fostered the creation of the Quantum Internet Alliance (QIA) with €24 million in funding to build “a global quantum internet made in Europe”.

Asia Test beds

China leads in quantum satellite communications, linking satellite QKD with terrestrial fiber networks. A team of researchers in China has demonstrated an integrated space-to-ground quantum communication network that combines a fiber network of more than 700 fiber links and two high-speed satellite-to-ground free-space links. Projects are also underway in Japan, India, Singapore, and South Korea.

Wrapping Up

Challenges remain in the march toward a fully functioning quantum network. Reliable quantum memories are one. Another important missing piece is the ability to extend the reach of a quantum link to arbitrarily long distances, using quantum repeaters. Quantum states cannot be simply copied and regurgitated, as is done with classical information. Quantum nodes will need sophisticated quantum logic gates to ensure that entanglement is preserved in the face of losses from interaction with the environment. Commercial projects and technologies are beginning to be introduced into the market. We can expect quantum networking to become mainstream quickly.

Let us know what you think. Please share your thoughts via the Comments section for this post or open a new SubStack chat thread … and please forward this post to your friends and colleagues. See you next time!