Quantum Computing Companies

Oxford Quantum Circuits (OQC)

(This profile is one entry in my 2025 series on quantum hardware roadmaps and CRQC risk. For the cross‑vendor overview, filters, and links to all companies, see Quantum Hardware Companies and Roadmaps Comparison 2025.)

Introduction

Oxford Quantum Circuits (OQC) is a UK-based quantum computing company founded in 2017 as a spin-out from the University of Oxford. It has emerged as a leading hardware developer focused on building commercially useful quantum computers for real-world applications. OQC was the first European provider of Quantum-Compute-as-a-Service (QCaaS), delivering enterprise-grade quantum systems via the cloud and even deploying them in standard commercial data centers.

At the core of OQC’s technology is its proprietary “Coaxmon” architecture – a three-dimensional superconducting qubit design (a variant of the transmon) that places qubit components on opposite sides of a substrate to simplify wiring and improve coherence. By leveraging this 3D coaxial approach, OQC aims to overcome some scaling limitations of planar superconducting circuits while maintaining compatibility with proven microwave control techniques.

Overall, OQC’s mission is to pioneer a quantum-accelerated world by building the first enterprise-ready fault-tolerant quantum computers, focusing on hardware innovations that bring quantum advantage closer to reality.

Milestones & Roadmap

OQC’s development timeline features several notable milestones in hardware and a published roadmap outlining its ambitious path to fault-tolerant quantum computing:

2017: Oxford Quantum Circuits is founded as a university spin-out, beginning development of its Coaxmon-based superconducting qubit technology.

2021-22: Debuted an 8-qubit prototype “Lucy”, which by February 2022 became the first European quantum processor available on AWS Braket (Amazon’s quantum cloud platform). This marked the world’s first public quantum computer launch outside North America, positioning OQC as a pioneer in cloud quantum access.

2023: Deployed the 32-qubit “Toshiko” Gen-1 system as the world’s first enterprise-ready quantum platform in commercial data centers (with installations in London and Tokyo)oqc.techoqc.tech. Toshiko uses OQC’s Coaxmon qubits and achieved two-qubit gate fidelities around 99%, demonstrating competitive performance at scale. OQC’s devices became accessible via its cloud service and direct colocation, enabling businesses to experiment with quantum computing using standard tools like QIR, OpenQASM, and OpenPulse.

2024: Secured an additional $100 million Series B funding (following a $40 million Series A in 2022) to accelerate hardware development. OQC at this point remained the only organization to successfully deploy quantum computers directly into third-party commercial data centers, underscoring its focus on operationalizing quantum hardware.

2025: Published a comprehensive technical roadmap (2025-2034) targeting ~200 logical qubits by 2028 and 50,000 logical qubits by 2034, far beyond any other publicly disclosed quantum roadmap. This roadmap formalizes OQC’s transition from the “physical qubit era” to the “logical qubit era,” emphasizing error-corrected logical qubit counts over raw physical qubits. Key planned stages include:

  • OQC Genesis (~2026): A “KiloQuOp” scale system (~1,000 quantum operations per second) featuring 16 logical qubits with built-in error suppression. Genesis will use OQC’s patented dual-rail “Dimon” qubit design to encode a logical qubit in the footprint of a single physical qubit, inaugurating OQC’s logical-qubit era. It’s slated to be commercially available in 2026 as the first device with hardware-level error detection at the qubit level.
  • OQC Titan (2028): A “MegaQuOp” system (~1,000,000 ops) featuring 200 logical qubits implemented within ~2,000 physical qubit lattice sites on a single 100 mm wafer. Titan is projected to run at a 1 MHz quantum operation clock speed and achieve quantum advantage on certain problems, outperforming classical supercomputers in domains like financial optimization and defense-related computations. By around 2028, Titan is expected to provide practical commercial advantage to users, aligning with the UK’s national goal of a “Million-QOp” quantum computer by that time.
  • OQC Athena (early 2030s): A “GigaQuOp” next-generation system scaling to 5,000 logical qubits on a 200 mm wafer. With an error-corrected logical error rate targeted around 10^−9, Athena would be capable of billions of quantum operations, opening up transformative applications in quantum chemistry, materials simulation, and advanced finance (e.g. complex derivative pricing and risk analysis). Athena’s design includes a wafer-scale QPU (fabricated on 200 mm wafers) and modular system integration to support tens of thousands of physical qubits, pushing OQC’s hardware into true foundry-level fabrication.
  • OQC Atlas (2034): The end-goal “TeraQuOp” system delivering ≈50,000 logical qubits with a logical error rate on the order of 10-12. Atlas is envisioned as a fault-tolerant quantum computer constructed from two coupled 300 mm wafers (total ~1 million physical qubits). This ambitious machine would redefine what’s computationally possible – enabling trillions of operations for applications like drug discovery, large-scale quantum chemistry, AI, and breaking cryptographic codes. OQC’s roadmap sets Atlas for 2034, signaling a bold vision to achieve a fully error-corrected, cryptographically relevant quantum computer within nine years.

Each successive generation in OQC’s roadmap aims to dramatically increase the logical qubit count and decrease error rates through hardware innovation, rather than brute-force scaling of physical qubits. Notably, 200 logical qubits by 2028 (Titan) is expected to unlock valuable capabilities such as fraud detection, cyber-threat analysis and optimization tasks, while 50,000 logical qubits by 2034 (Atlas) would push into regimes of decryption and quantum-enabled discovery previously thought to be decades away. OQC’s timeline is aggressive, but it highlights a clear trajectory from today’s noisy devices to tomorrow’s fault-tolerant quantum computers.

Focus on Fault Tolerance

From its research through its roadmap, OQC places a strong focus on fault tolerance and error correction at the hardware level. The company’s strategy hinges on reducing the overhead required for quantum error correction by engineering more error-resilient qubits from the start. A centerpiece of this approach is OQC’s dual-rail “Dimon” qubit technology, a patented design that encodes a qubit’s state across two modes (or “rails”) to allow error detection at the individual qubit level. In new results published in 2025, OQC demonstrated that this dual-rail Dimon approach can produce “reproducible error-suppressed qubits,” dramatically slashing the number of physical qubits needed per logical qubit. In fact, OQC claims its architecture achieves a physical-to-logical qubit ratio about 10× more efficient than conventional quantum error correction codes (which often require dozens or hundreds of physical qubits per logical qubit). By catching and suppressing errors in hardware (within each qubit’s design), far fewer qubits are needed to build robust logical qubits – a crucial advantage on the road to fault tolerance.

This hardware-efficient error correction is complemented by more traditional QEC codes applied at the system level. For example, the upcoming Titan system will incorporate concatenated error-correcting codes on top of the Dimon qubits, enabling it to sustain millions of quantum operations per algorithm run without decoherence accumulating. Logical error rates are projected to improve from ~10-3 in the 16-qubit Genesis, to ~10-6 in the 200-qubit Titan, and all the way to 10-12 in the 50k-qubit Atlas machine. Achieving such ultra-low error rates would fulfill the requirements for fault-tolerant quantum computing, wherein algorithms of essentially arbitrary length can run reliably. OQC’s recent integration of Riverlane’s Deltaflow real-time QEC system into its hardware testbed underscores this focus – it’s the first deployment of dedicated QEC technology in a commercial quantum data center in the UK. This joint effort (part of the UK’s DECIDE project) is intended to validate live quantum error correction on OQC’s superconducting qubits and is a critical step toward scaling to logical qubits “kept alive” for millions of operations.

By investing in both novel qubit architectures and an integrated error-correction stack, OQC is explicitly steering toward fault tolerance. The company’s philosophy is that true commercial value in quantum computing will only be realized once error-corrected logical qubits are available – hence the emphasis on the “logical era” in its roadmap. If OQC’s approach works as hoped, it could significantly bring forward the timeline for useful quantum computers by reducing the massive hardware overhead that has traditionally been assumed necessary for QEC. This could make scalable quantum computing feasible with thousands (or even hundreds) of physical qubits, instead of millions, potentially leapfrogging competitors in the race to build the first fault-tolerant machine.

CRQC Implications

CRQC (“Cryptographically Relevant Quantum Computing”) refers to a quantum computer powerful enough to break modern cryptographic algorithms like RSA and ECC. OQC’s roadmap has clear implications in this realm: by the Atlas stage (~50,000 logical qubits, expected by 2034), OQC explicitly anticipates that decryption of classical encryption schemes will be within reach. At 50k logical qubits, a machine could run Shor’s algorithm to factor RSA-2048 or solve discrete logarithms for elliptic-curve cryptography, thereby undermining the security of current public-key encryption protocols. In OQC’s vision, Atlas would mark the point where quantum computers transition from being mostly scientific curiosities to being direct threats and tools in cybersecurity – enabling, for example, the decryption of RSA-encrypted data and other cryptanalytic tasks that are infeasible for classical computers. This is why applications like “decryption” and “cybersecurity” appear among Atlas’s target use cases in OQC’s roadmap documentation.

Even before reaching full CRQC capability, OQC’s interim milestones carry security implications. By 2028, the 200-logical-qubit Titan is expected to have “significant implications for cybersecurity” – not necessarily by cracking RSA directly, but by enabling advanced algorithms for things like quantum-enhanced cryptanalysis, secure communications, and threat detection. OQC’s leadership has underscored the urgency for organizations in finance and national security to prepare for a quantum-transformed world. The roadmap’s unveiling was a call-to-action, noting that practical quantum computing may arrive sooner than many expect and that the “moment when quantum computing begins to transform people’s lives is closer than many realize”. In concrete terms, if OQC (or any competitor) stays on track, the late 2020s to early 2030s could see the advent of cryptographically relevant quantum machines. This adds pressure on the cybersecurity world to accelerate post-quantum cryptography defenses in parallel. In summary, OQC’s technical progress, if it continues as planned, will play directly into the timeline for CRQC, with Atlas representing a potential quantum tipping point for modern encryption. Governments and industries are taking note, as a 50,000-logical-qubit device would firmly usher in the post-quantum crypto era.

Modality & Strengths/Trade-offs

OQC firmly backs superconducting qubits (transmon-style qubits in a dilution refrigerator) as its modality of choice, leveraging this platform’s well-understood physics and fast operation speeds. Its unique contribution is the Coaxmon design – a patented 3D packaging of a transmon qubit. In a coaxmon, the qubit sits on one side of a sapphire substrate while the readout resonator and control wiring interface from the opposite side, with a via through the substrate to couple them. This geometry eliminates the crowded in-plane wiring of conventional 2D transmons and reduces cross-talk, potentially improving coherence times and gate uniformity across the chip. The coaxial approach also makes it easier to scale to larger chips (and wafers) since control lines can be fed from below without obstructing the qubit plane. In terms of performance, OQC has reported one of the fastest high-fidelity 2-qubit gates in the industry: a controlled-Z gate executed in only ~25 ns with 99.8% fidelity. Such speedy gates (an order of magnitude faster than many other superconducting qubit architectures) mean fewer errors per operation and are “among the fastest and most accurate gates ever recorded” for any quantum technology. Fast gate speeds are a significant strength of OQC’s modality, as they increase the feasible circuit depth before decoherence sets in, thus expanding the algorithmic complexity that can be handled on near-term devices.

Another strength of OQC’s hardware is the multi-mode “Dimon” qubit capability, which effectively embeds a small error-detecting code within each physical qubit. By using two resonant modes to represent the |0⟩ and |1⟩ states (the dual-rail encoding), a Dimon qubit can detect certain spontaneous error events (like a leakage or photon loss in one rail) and suppress them, something a standard transmon cannot do. This gives OQC’s qubits an intrinsic resilience and drastically lowers the overhead when building large error-corrected circuits – a 10× resource advantage as OQC claims, compared to other superconducting approaches that might need 100+ physical qubits per logical qubit. In practice, this means OQC can aim for fault-tolerance with thousands of physical qubits rather than millions, making its roadmap more plausible if the technology delivers as expected.

Like all superconducting qubit platforms, however, OQC’s modality faces trade-offs. The reliance on cryogenic infrastructure (dilution refrigerators at ~10 mK) and complex room-temperature control electronics is a challenge as systems grow. OQC’s plan to fabricate on 200 mm and 300 mm wafers and even couple multiple wafers introduces engineering hurdles in maintaining uniform qubit quality, calibrating large arrays, and managing heat load and crosstalk in a fridge environment. While the coaxmon design eases wiring congestion, the sheer scale of an Atlas-class system (1,000,000 physical lattice sites) pushes the limits of today’s fabrication and cryo-cooling technology. OQC’s bet on superconducting qubits reflects a belief that these challenges are surmountable in exchange for the advantages of maturity and speed that superconducting circuits offer. The platform has a long track record (used by Google, IBM, etc.), a rich ecosystem of control software and electronics, and known pathways to improve coherence (materials purification, 3D integration, etc.). OQC is enhancing those strengths with its own innovations – aiming to outperform not only other superconducting efforts but also alternate modalities like ion traps or photonics in the race to useful scale. OQC asserts that its hardware will deliver higher quality results faster than competing systems by requiring fewer qubits and achieving faster gates. If Coaxmon and Dimon can be realized as designed, OQC’s modality could offer a rare combination: the speed and connectivity of superconducting circuits with the error robustness approaching that of architecturally redundant systems. The trade-off, of course, is that OQC must solve some of the hardest physics and engineering problems of quantum scaling concurrently, within a startup-sized organization. The next section examines how they have managed so far.

Track Record

OQC’s track record to date shows a pattern of consistent hardware progress and industry firsts in quantum computing, lending credibility to its ambitious roadmap. One of its early achievements was the launch of the “Lucy” 8-qubit quantum processor on AWS Braket in 2022, which made OQC the first European quantum hardware provider on Amazon’s cloud. Lucy operated as a stable cloud-accessible system with over 98% uptime in its first two years, showcasing OQC’s ability to deliver reliable quantum hardware-as-a-service. This was a significant milestone not just for OQC but for the European quantum community, as it put Europe’s first quantum machine alongside offerings from U.S.-based firms on a global platform.

Building on Lucy, OQC developed “Toshiko” Gen 1, a 32-qubit device deployed in 2023. Toshiko represented the world’s first quantum computer installed directly in a commercial data center (OQC partnered with data center providers to host their cryogenic system on-site). Enterprise clients gained access through OQC’s cloud and via private connections, demonstrating OQC’s commitment to making quantum computing enterprise-ready and accessible. Toshiko’s performance was competitive: median single-qubit gate fidelity ~99.9% and two-qubit fidelity ~99%, with respectable qubit coherence times (~100 µs). These specs, published by OQC, indicate a quality on par with other leading superconducting efforts worldwide. The successful deployment of a 32-qubit system with these characteristics gave OQC a credible platform to iterate upon. Indeed, the OQC Genesis (2026) device will be an evolution of the Toshiko architecture augmented with dual-rail Dimon qubits, showing a clear lineage from the current generation to the fault-tolerant designs.

OQC’s partnerships and ecosystem involvement further bolster its track record. The company has collaborated with classical computing and cryptography communities, for instance by showcasing its hardware at events like DEFCON’s Quantum Village (where Lucy was used in quantum capture-the-flag challenges) and by working with firms like Classiq and Riverlane on software integration. The Riverlane collaboration in 2025, integrating a real-time QEC decoder into OQC’s system, was the first of its kind in the UK. Additionally, OQC’s leadership and advisors include notable figures – for example, former GCHQ head Sir Jeremy Fleming joined OQC’s board in 2025, reflecting the strategic importance of its technology. In terms of funding, OQC’s ability to attract large investments is noteworthy: its $100 million Series B in 2024 was one of the largest quantum computing funding rounds in Europe, providing runway to execute its roadmap. The company’s investors and partners (which span government initiatives, cloud providers, and industry end-users) indicate that OQC is viewed as a key player in the UK and EU quantum programs, often cited as a national champion in quantum hardware.

Crucially, OQC has demonstrated an ability to deliver on its hardware goals so far. Each generation (from an initial 4-qubit prototype named “Sophia”, to Lucy’s 8 qubits, to Toshiko’s 32 qubits) has markedly increased qubit count and functionality, roughly on schedule. While many quantum startups remain in R&D mode, OQC already offers commercial cloud access to its quantum processors and has real customers experimenting on its machines. This practical track record – operating quantum systems in the field – differentiates OQC and provides valuable operational experience as they scale to larger devices. In summary, OQC’s achievements to date (first QCaaS in Europe, multi-qubit deployments with high fidelities, integration of new technologies like Dimon qubits and QEC stacks) give weight to its roadmap projections. The company has shown a pattern of steady progress, though the most challenging steps undoubtedly lie ahead.

Challenges

Despite OQC’s impressive progress, significant challenges remain on the road to its fault-tolerant goals:

Scaling Hardware Complexity: As OQC moves toward chips with thousands then millions of qubits, engineering hurdles multiply. Maintaining ultra-low error rates while cooling and controlling so many qubits in a dilution refrigerator is a non-trivial problem. The team must devise solutions for signal routing (potentially using 3D integration and cryoelectronics) to avoid a wiring bottleneck as devices grow larger. Each increase in wafer size (100 mm → 200 mm → 300 mm) introduces fabrication challenges around uniformity, yield, and packaging of the quantum chips.

Error Correction Overhead and Speed: Even with the Dimon qubit efficiency, implementing full real-time error correction on thousands of qubits will push the limits of classical co-processing and feedback. OQC will need sophisticated classical hardware (FPGAs or custom ASICs) to decode and correct errors on-the-fly without slowing down the quantum processor. Ensuring the cycle time of error detection/correction is fast enough (on the order of microseconds) is essential so that logical qubits can be sustained. This requires tight integration of hardware and software and is an area of active risk and research.

Unknown Physics at Scale: OQC is entering regimes no superconducting qubit system has been before. Unforeseen issues (“unknown unknowns”) could emerge when scaling to wafer-scale quantum processors. These could include new noise sources, crosstalk patterns, heating effects, or materials defects that were negligible at smaller scales. For instance, a 1,000-qubit chip might encounter anomalous error correlations or drifts that current 50-qubit experiments have not revealed. OQC’s aggressive roadmap leaves little room for unexpected setbacks that require fundamental R&D breakthroughs.

Timeline and Execution Risk: Achieving the stated milestones by 2028 and 2034 is extremely ambitious. OQC operates in a competitive landscape where giants like IBM, Google, and IonQ are also racing toward fault tolerance (albeit with differing approaches). There is a risk that delays in any single milestone could cascade – for example, if Genesis in 2026 underperforms or is late, it compresses the time to develop Titan by 2028. As a startup, OQC must scale its team and infrastructure rapidly to meet these timelines, which is a managerial and logistical challenge in itself. The success of the plan depends on flawless execution across multiple fronts (hardware, firmware, fabrication, partnerships) with relatively limited resources compared to big-tech competitors.

Competition and Market Adoption: While OQC focuses on hardware, the ultimate impact also depends on the software and use-cases. They must ensure a robust software stack and developer ecosystem so that customers can actually leverage 200 logical qubits by 2028. If, for instance, competing modalities (like neutral-atom or photonic qubits) achieve useful scale sooner, OQC will face pressure. At 50,000 logical qubits, cryptographic breaking becomes possible – but governments might already have transitioned to quantum-safe algorithms by then, potentially diminishing that particular “killer application.” OQC will need to keep demonstrating intermediate quantum advantages (e.g. in optimization or simulation) to maintain commercial relevance on the way to the end goal.

In summary, OQC confronts the same core challenges as the rest of the industry – reducing errors, scaling qubit count, and integrating complex systems – but at an even more extreme scale given its bold roadmap. The technical challenges of large-scale superconducting quantum machines (thermal management, crosstalk, chip fabrication, fast control electronics) are daunting. That said, OQC’s focused innovations (like coaxmon and Dimon) are intended to address some of these pain points. The next few years will be critical to see if these solutions pan out in practice.

Marin Ivezic

I am the Founder of Applied Quantum (AppliedQuantum.com), a research-driven consulting firm empowering organizations to seize quantum opportunities and proactively defend against quantum threats. A former quantum entrepreneur, I’ve previously served as a Fortune Global 500 CISO, CTO, Big 4 partner, and leader at Accenture and IBM. Throughout my career, I’ve specialized in managing emerging tech risks, building and leading innovation labs focused on quantum security, AI security, and cyber-kinetic risks for global corporations, governments, and defense agencies. I regularly share insights on quantum technologies and emerging-tech cybersecurity at PostQuantum.com.
Share via
Copy link
Powered by Social Snap