Quantum Computing – Hardware Designs – Part 1 of a 2 Part Series
An overview of Majorana, Willow and Ocelot (in easy to understand terminology) :-)
In this post, the first of 2, we explore hardware designs being developed and commercialized to bring Quantum Computing to the masses! 😊 In late 2024 and early 2025, Google, Microsoft, and Amazon each unveiled proprietary quantum chips, crystallizing a new stage in the race for quantum computing dominance.
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!
Note: Chinese researchers are also building quantum computing chips – example: Zuchongzhi 3.0. We are not covering those providers in this article.
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.
Watch for Part 2 of this series on Quantum Computing Hardware Designs – coming soon!
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!