A Brief History of Quantum Computing

Billions Invested, Big Bets Placed, and a Race That’s Only Just Beginning

From the first quantum demonstrations in university labs to Google’s landmark supremacy experiments, from early-stage SPAC listings to billion-dollar government programs — the story of how an entire industry took shape around one of the most ambitious engineering challenges in history.

2000 – 2026

The Greatest Technology Race Nobody Fully Understands

Quantum computing is the field where physics, engineering, and capital markets intersect in ways that rarely follow a straight line. It is a discipline where scientists debate definitions, companies stake decades on unproven approaches, and governments write nine-figure cheques based on projections that may or may not hold. And yet, beneath the noise, something real and consequential is happening.

Since the year 2000, researchers, companies, and governments have been building toward computers that compute in a fundamentally different way — leveraging the properties of quantum mechanics to solve problems that classical machines simply cannot. The journey has been marked by genuine breakthroughs, significant setbacks, vigorous scientific debate, and investments on a scale that reflects how seriously the world has come to take the promise.

This is the story of that journey — the people who drove it, the technology bets that won and lost, and the landscape as it stands today.

✦ ✦ ✦

Part I: The Academic Foundations (2000 – 2006)

From Theory to Early Experiment

In the year 2000, quantum computing was primarily an academic pursuit. IBM maintained a research laboratory in Yorktown Heights, New York, where scientists were demonstrating that quantum algorithms could run on a handful of qubits built from nuclear magnetic resonance (NMR) techniques — molecules in a magnetic field whose nuclear spins could be controlled and measured with remarkable precision. It was scientifically impressive, and it established key principles. It was also clear that NMR would not scale to the system sizes needed for practically useful computation.

The theoretical case for quantum computing had already been established. Peter Shor had published his factoring algorithm in 1994, demonstrating that a sufficiently powerful quantum computer could break widely used encryption schemes. Lov Grover’s 1996 search algorithm showed quantum advantages for database problems. And Michael Nielsen and Isaac Chuang’s 2000 textbook, Quantum Computation and Quantum Information, provided the field with its definitive educational foundation — a book that remains the standard reference a quarter century later.

The central engineering challenge was hardware. Qubits — the quantum equivalent of classical bits — had to be maintained in a coherent quantum state, which meant keeping them isolated from environmental disturbances at temperatures close to absolute zero. The moment a qubit interacted with its environment, a process called decoherence, its quantum information was lost. Building systems that could control many such qubits simultaneously, with low error rates, was an immense materials, fabrication, and control engineering challenge.

IBM was methodically working through this with superconducting qubit designs. Others were exploring alternative physical platforms. And from Burnaby, British Columbia, a company was about to take a very different approach entirely.

“The name of the game from the outset was to make a functional computer. Then we could probe it to see where it was operating correctly.” — D-Wave, on their engineering philosophy

D-Wave Systems: A Different Kind of Quantum Machine

D-Wave Systems, founded in 1999 by Geordie Rose and colleagues at the University of British Columbia, chose a path that diverged sharply from the academic mainstream. Rather than pursuing a universal gate-based quantum computer capable of running any algorithm, D-Wave focused on a specialized technique called quantum annealing — an approach designed specifically for optimization problems. The idea was to use quantum tunneling to help find the lowest-energy solution to a complex landscape of possibilities, a class of problem that arises constantly in logistics, finance, chemistry, and materials science.

D-Wave built their qubits from superconducting loops of niobium, cooled to near absolute zero, and pursued a strategy of rapid hardware development — prioritizing demonstrated functionality over the peer-reviewed incremental progress favoured by academic labs. This approach generated significant controversy among quantum physicists, many of whom argued that the company should establish the quantum properties of their system before making commercial claims. D-Wave maintained that demonstrated performance on real problems was a valid measure of progress.

On February 13, 2007, D-Wave demonstrated their Orion system at the Computer History Museum in Mountain View, California — a 16-qubit adiabatic quantum processor solving optimization problems in a live public setting. The event attracted both significant media attention and sustained scientific scrutiny. Questions about whether D-Wave’s machines were genuinely harnessing quantum effects for computational advantage — rather than simply mimicking classical annealing — would continue for years.

What was not in question was the commercial interest. In 2011, D-Wave announced the D-Wave One, an integrated 128-qubit system sold commercially for approximately $10 million. Lockheed Martin became the first customer. The D-Wave Two, a 512-qubit successor, was jointly acquired by NASA and Google in 2013 for the Quantum Artificial Intelligence Lab at NASA’s Ames Research Center. Regardless of where one stood on the technical debates, D-Wave had achieved something no other quantum company had: a paying customer base.

✦ ✦ ✦

Part II: The Cloud Opens and the Startups Arrive (2016 – 2019)

IBM Puts Quantum on the Internet

While D-Wave pursued its specialized path, IBM continued developing universal gate-based quantum computers using superconducting transmon qubits. Their steady progress culminated in a landmark move in 2016: IBM made a quantum computer available on the cloud. The IBM Quantum Experience gave any researcher, student, or developer access to a 5-qubit quantum processor via a web browser.

The decision to democratize access was strategically significant. IBM built an enormous user community, catalysed quantum programming education globally, and established a strong position as the public face of accessible quantum computing. They followed with a series of hardware milestones — 16 qubits in 2017, a prototype 50-qubit system later that year, and in 2019 the IBM Q System One: an integrated 20-qubit commercial system, housed in a striking industrial design intended to signal quantum’s readiness for enterprise environments.

IBM also introduced the concept of Quantum Volume as a holistic hardware benchmark — a single metric factoring in qubit count, error rates, connectivity, and coherence time. This was a meaningful contribution to the field: raw qubit count, they argued, was a poor proxy for actual computational capability, and the industry needed better ways to compare systems across different architectures.

A New Generation of Companies

From 2015 to 2018, a wave of quantum computing startups emerged, each making a different architectural bet. The diversity of approaches reflected the field’s fundamental openness: there was no established consensus on which physical platform would ultimately win.

Chad Rigetti, who had completed his doctorate at Yale and worked at IBM, founded Rigetti Computing in 2013. His thesis was that building a better superconducting quantum computer required controlling the full manufacturing stack — including operating your own fabrication facility. Rigetti built a quantum chip fab in Fremont, California, and pursued tight integration between quantum and classical computing from the outset. The company raised substantial venture capital — $64 million in 2017, $71 million in 2019 — and attracted serious talent from academia and industry alike.

IonQ launched in 2015 out of the University of Maryland, founded by Chris Monroe and Jungsang Kim, two of the world’s leading experts in trapped-ion physics. Their approach used individual ytterbium atoms suspended in electromagnetic traps and manipulated with precisely tuned lasers. Trapped ions had a significant advantage over superconducting qubits in coherence time — they could maintain quantum states far longer, enabling more complex computations before errors accumulated. The tradeoff was speed: individual gate operations on ions were substantially slower. IonQ’s founders believed that quality of gates would ultimately matter more than raw clock speed, and built accordingly.

In the same period, Jeremy O’Brien — an Australian physicist who had led pioneering work in optical quantum computing at the University of Bristol — co-founded PsiQuantum in 2016 with colleagues Terry Rudolph, Peter Shadbolt, and Mark Thompson. PsiQuantum’s approach was photonic: using individual photons in silicon chips as qubits. Their central argument was that useful quantum computing would require millions of qubits and that the only credible path to manufacturing at that scale was through existing semiconductor foundries. PsiQuantum therefore focused not on building small demonstrators, but on designing photonic quantum chips manufacturable in high-volume silicon fabs from the outset — a strategy that made them unusual in an industry otherwise building from small systems up.

Microsoft, drawing on years of theoretical research, had made a major commitment to topological qubits — a fundamentally different approach based on exotic quasiparticles called Majorana fermions. The theoretical appeal was substantial: topological qubits would store quantum information in a way that is inherently protected from local noise, potentially reducing the error-correction overhead that was expected to consume the vast majority of qubits in any fault-tolerant system. In 2017, Microsoft significantly expanded its quantum effort, establishing research groups across multiple continents and committing to topological qubits as its primary hardware path.

Also entering the picture was Honeywell. The industrial conglomerate had quietly built a world-class quantum research team, and their trapped-ion systems — operated with precision more commonly associated with atomic clocks than computers — were achieving some of the highest gate fidelities in the field.

October 2019: Quantum Supremacy and Its Debate

On October 23, 2019, Google published a landmark paper in Nature describing an experiment performed on their 53-qubit Sycamore processor. The chip had completed a specific random circuit sampling task in 200 seconds — a task that, Google estimated, would require approximately 10,000 years on the most powerful available classical supercomputer.

Google described this as ‘quantum supremacy’: the first demonstration that a quantum processor could perform a well-defined computation faster than any classical machine could feasibly replicate. CEO Sundar Pichai called it the ‘hello world’ moment of quantum computing, drawing an analogy to the Wright brothers’ first flight — a milestone that proved a concept without yet delivering commercial utility.

The announcement generated immediate scientific debate. IBM’s research team published a response arguing that their Summit supercomputer, with optimized classical simulation strategies, could perform the same calculation in approximately 2.5 days rather than 10,000 years — making the practical gap between quantum and classical performance considerably smaller than Google’s paper implied. IBM also raised concerns about the terminology, arguing that ‘quantum supremacy’ overstated the significance of a benchmark designed specifically to be hard for classical simulators, rather than one with real-world applications.

The scientific community engaged seriously with both positions. The underlying achievement — a 53-qubit processor operating with sufficient fidelity to produce results that pushed the boundaries of classical simulation — was broadly recognized as significant. The precise interpretation of that result, and whether ‘supremacy’ was the right framing, remained actively debated. What was not in doubt was that quantum computing had become a major story, and that governments, investors, and enterprise technology teams were paying close attention.

“Quantum supremacy does not equal quantum value.” — Brian Hopkins, Forrester Research, October 2019

✦ ✦ ✦

Part III: The Capital Markets Chapter (2020 – 2023)

Public Markets Discover Quantum

The combination of Google’s 2019 headline and low interest rate environments created favorable conditions for quantum computing companies to access public capital. The vehicle of choice was the SPAC — Special Purpose Acquisition Company — a mechanism that allowed companies to list on public markets through a merger with a blank-check acquisition entity, bypassing some of the traditional IPO process.

IonQ went public via SPAC in 2021 at a valuation of approximately $2 billion, becoming the first pure-play quantum hardware company on a public exchange. Rigetti Computing followed in March 2022 at a $1.5 billion valuation. D-Wave listed shortly after at $1.6 billion. All three were pre-revenue in any meaningful commercial sense — IonQ had generated $2 million in 2021, Rigetti approximately $13 million against an $18 million forecast, D-Wave around $1.7 million per quarter — but investor appetite for deep technology with long-horizon returns was strong.

The macroeconomic environment shifted sharply through 2022. Rising interest rates reduced appetite for speculative long-duration technology investments, and all three quantum stocks fell significantly from their peak valuations. Rigetti, which had opened at $9.75 per share, reached a low of $0.38 in May 2023. The company faced potential Nasdaq delisting, restructured its leadership team with founder Chad Rigetti departing as CEO in December 2022, and reduced its workforce by 28%. D-Wave traded below $1. IonQ maintained somewhat more resilience but remained well below its listing valuation.

These were difficult years for the sector, but they also reflected a correction that most observers considered structurally necessary. Public markets had priced quantum computing companies on projections that assumed faster commercial traction than the hardware had yet delivered. The period of adjustment was painful for investors; it did not meaningfully slow technical progress at the companies involved.

Quantinuum: Quality Over Quantity

Honeywell’s quantum effort had been maturing quietly through this period. In 2020, Honeywell Quantum Solutions announced that their trapped-ion H1 system had achieved a record Quantum Volume score — IBM’s own benchmark metric — a milestone that signalled the competitiveness of ion-trap hardware on the fidelity dimensions that IBM had argued mattered most. The claim prompted a spirited public exchange between the two companies’ technical teams about the precise interpretation of the benchmark.

In 2021, Honeywell spun out its quantum computing division and merged it with Cambridge Quantum Computing — a UK-based quantum software and algorithms company — to form Quantinuum. The combined entity pursued what it described as a ‘full-stack’ strategy: world-class trapped-ion hardware integrated with quantum software, chemistry simulation tools, and cybersecurity applications. Their H-series processors continued to advance gate fidelity systematically, and the company began reporting commercial revenue from enterprise customers in chemistry and finance. Quantinuum was building a business model, not just hardware.

✦ ✦ ✦

Part IV: The Qubit Race Intensifies (2022 – 2024)

IBM’s Public Roadmap

IBM had, since 2020, distinguished itself by committing publicly to a detailed multi-year hardware roadmap with specific qubit targets: 65 qubits in 2020, 127 in 2021, 433 in 2022, and 1,121 in 2023, leading toward a vision of 100,000-qubit quantum-centric supercomputers by 2033. Publishing forward commitments in a field this technically uncertain was a meaningful act of accountability, and IBM largely delivered against them.

Their November 2022 Eagle successor, Osprey, reached 433 qubits. In December 2023 at their annual Quantum Summit, IBM unveiled Condor: 1,121 superconducting qubits on a single chip, the first to exceed 1,000. They simultaneously announced Heron — a smaller, 133-qubit processor engineered for higher fidelity and optimized for modular connection with other chips. The modular architecture reflected a recognition that simply increasing qubit count on a single chip was reaching engineering limits, and that future scaling would require connecting many smaller processors.

IBM also published research that year demonstrating what they called ‘quantum utility’ — evidence that their 127-qubit Eagle processor could perform simulations of condensed matter physics that were genuinely challenging for the best available classical approximation methods. It was a cautious claim, carefully bounded, but it pointed toward a new framing: not supremacy over classical computers in general, but quantum processors becoming useful tools for specific scientific problems.

Atom Computing: A Milestone from an Unexpected Direction

On October 24, 2023 — just weeks before IBM’s planned Condor announcement — Atom Computing revealed that their second-generation neutral atom system had been loaded with 1,180 qubits, making it the first gate-based quantum system to exceed 1,000 qubits.

Atom Computing, founded in 2018 and having raised approximately $91 million total, used a different physical platform entirely: neutral atoms of ytterbium, held in place and manipulated by precisely tuned laser beams in a vacuum chamber. The atoms were arranged in a 35-by-35 optical tweezer array. Where IBM’s superconducting qubits were fabricated on chips through semiconductor processes, Atom’s qubits were literally individual atoms — positioned, controlled, and measured with lasers.

The neutral atom approach had notable properties. Atom’s qubits demonstrated coherence times of 40 seconds — orders of magnitude longer than superconducting qubits, which typically maintained coherence for microseconds. The atoms were also naturally identical, eliminating fabrication variability. The chief limitation was gate speed: neutral atom operations were roughly 100 to 1,000 times slower than superconducting gates. Whether the coherence time advantage outweighed the speed disadvantage depended heavily on the application.

Atom Computing’s announcement illustrated a broader point about the state of the field: the leaderboard on any given metric could change rapidly, and from directions the incumbents hadn’t anticipated. IBM acknowledged the milestone while noting that qubit count was one of several relevant dimensions of system performance. The race was genuinely multi-dimensional.

Neutral atom approaches demonstrated that rapid scaling was achievable through a fundamentally different set of engineering tradeoffs than superconducting systems — expanding the range of credible paths to useful quantum computation.

Neutral Atoms and Error Correction: QuEra’s Contribution

QuEra Computing, a Boston-based company spun out of Harvard University in 2019 and backed by Google and SoftBank, was also advancing neutral atom technology. In 2023, researchers at QuEra working with teams at Harvard, MIT, and the University of Maryland published a landmark result in Nature: a demonstration of 48 logical qubits, operating with error rates below those of the underlying physical qubits.

This was a significant theoretical milestone. Logical qubits — multiple physical qubits combined through error-correcting codes into a single reliable computational unit — had been the theoretical goal of the field for decades. Demonstrating that encoding qubits logically actually improved their performance, rather than simply adding overhead, was an important proof of concept. The field had moved from asking whether fault-tolerant quantum computation was theoretically possible to asking when it would become practically achievable.

Microsoft: A Long and Contested Road

Microsoft’s topological qubit program had an eventful several years. A 2018 paper by a Microsoft-affiliated team in Nature reported strong evidence for Majorana zero modes — the quasiparticles at the heart of the topological approach. The paper was accompanied by cautious enthusiasm from parts of the scientific community and more pointed skepticism from others. In 2020, a note of concern was added to the paper; in 2021, it was retracted following an investigation that found the data had been incompletely presented. Microsoft acknowledged the finding and continued its research program under a more rigorous verification framework.

In 2023, the team published new experimental results in Physical Review B, describing a ‘topological gap protocol’ designed to more rigorously establish the presence of the topological phase. Scientific debate continued about whether the evidence was conclusive. Then, in February 2025, Microsoft unveiled Majorana 1 — a processor they described as the world’s first built on a topological core architecture — alongside a paper in Nature and a commitment to developing fault-tolerant systems based on this approach.

The scientific reception was mixed. Some researchers noted that the accompanying Nature paper, as reviewed, did not itself constitute proof of Majorana zero modes in the device. Others acknowledged the materials science progress as genuine, even while questioning the topological interpretation. Microsoft maintained confidence in their results. The debate underscored a recurring challenge in quantum computing: extraordinary claims require extraordinary evidence, and the experimental signatures of novel quantum phenomena are genuinely difficult to establish conclusively.

What was not in question was the scale of Microsoft’s commitment. They had invested heavily over nearly two decades in a high-risk, high-potential approach, and they continued to do so.

✦ ✦ ✦

Part V: The Era of Early Utility (2024 – 2026)

Jensen Huang’s Challenge

In January 2025, at the Consumer Electronics Show, NVIDIA CEO Jensen Huang offered a pointed assessment: useful quantum computing was likely 15 to 30 years away. The remark landed in an industry that had been building toward near-term commercial relevance, and it prompted a year of concerted effort to demonstrate results that would challenge that timeline.

The most technically significant response came from Google in December 2024, when they announced Willow: a 105-qubit superconducting processor that achieved a key theoretical milestone in error correction. As Google added more qubits to the system, the logical error rate went down rather than up — a property called ‘below threshold’ operation that is a prerequisite for scalable fault-tolerant quantum computing. No previous experimental system had demonstrated this behaviour clearly. Willow also completed a benchmark task in approximately five minutes that Google estimated would take classical supercomputers an astronomically long time to match. The caveats around benchmark interpretation remained, but the error correction result was considered technically robust by most observers.

IBM contributed a different kind of evidence. In September 2025, HSBC announced that using IBM’s Heron quantum processor, they had improved the accuracy of bond trading predictions by 34% compared to classical methods alone. It was a single result, carefully qualified, and HSBC was clear that quantum computing was not yet embedded in their production systems. But it was a credible, externally validated data point that a quantum processor had contributed to a commercially relevant financial calculation — something that had not existed before.

Quantinuum launched their Helios trapped-ion system in November 2025, claiming the highest gate accuracy of any commercial quantum system. The company reported that early access customers including SoftBank and JPMorgan Chase had used it for commercially relevant research in chemistry and finance. Helios was also designed to be programmable using familiar classical toolchains, reducing the barriers to adoption for software engineers without deep quantum expertise.

PsiQuantum: Engineering at Scale

PsiQuantum continued its distinctive strategy through this period: building photonic quantum chips in a semiconductor foundry and scaling toward the million-qubit systems that Jeremy O’Brien had argued from the outset were necessary for commercial utility. In September 2025, the company closed a $1 billion Series E round led by BlackRock, with participation from Temasek, Baillie Gifford, the Qatar Investment Authority, and NVIDIA’s venture arm — bringing total funding to $2.3 billion and valuation to $7 billion.

The capital was earmarked for breaking ground on utility-scale quantum computing sites in Brisbane, Australia — backed by $940 million in Australian and Queensland government funding — and in Chicago. PsiQuantum had also announced Omega, a photonic chipset manufactured at GlobalFoundries’ Fab 8 in New York, containing high-performance single-photon sources, superconducting detectors, and next-generation optical switches based on barium titanate. Manufacturing thousands of such wafers in a commercial foundry was itself a significant technical milestone. The stated goal remained a commercially useful, fault-tolerant machine operational by 2027.

Photonic Inc.: Silicon, Photons, and the Network-First Approach

Among the companies pursuing fault-tolerant quantum computing from first principles, Photonic Inc. — headquartered in Vancouver, British Columbia — represented an approach built around a specific and unusual insight: that the qubit platform and the networking platform should be the same technology.

Photonic was founded in 2016 by Dr. Stephanie Simmons, who held a Canada Research Chair in Silicon Quantum Technologies at Simon Fraser University. Her research had identified what she called the ‘T centre’ — a specific radiation-damage defect in silicon — as a uniquely promising qubit candidate. T centres have long-lived quantum spins, making them effective as qubits; they also have optical transitions at telecom O-band wavelengths, meaning they can emit and receive photons compatible with standard fiber-optic infrastructure. A qubit that could naturally interface with global telecommunications networks was, in Simmons’ framing, the missing component for scalable distributed quantum computing.

The architecture Photonic developed around this platform — called Entanglement First — was designed to scale in two dimensions simultaneously: upward, by integrating large numbers of silicon qubits on a single chip; and outward, by entangling multiple chip modules across telecom fiber links. The result was a system that could, in principle, grow by adding modules connected through existing fiber infrastructure, rather than requiring a single monolithic processor of ever-increasing scale. Their error-correction strategy relied on Quantum Low-Density Parity-Check codes, which Photonic argued could achieve fault tolerance with significantly fewer physical qubits than surface codes — the standard used by most competitors.

In November 2023, Photonic publicly unveiled both its architecture and a $100 million funding round that included Microsoft, the UK government’s National Security Strategic Investment Fund, and the British Columbia Investment Management Corporation. Simultaneously, the company announced a strategic collaboration with Microsoft to integrate Photonic’s hardware into Azure as the technology matured. In May 2024, Photonic demonstrated distributed entanglement between qubits in separate cryogenic modules connected by telecom fiber — a ‘teleported CNOT gate’ between physically separated silicon spin qubits, described as an industry first for this platform. A September 2025 paper in Nature Photonics described an electrically triggered spin-photon interface in silicon — advancing the integration of the platform toward manufacturable form factors.

In November 2025, DARPA selected Photonic as one of eleven companies for Stage B of its Quantum Benchmarking Initiative — a program aimed at verifying which architectural approaches have a credible path to utility-scale fault-tolerant quantum computing by 2033. Photonic’s stated internal target is more aggressive: a scalable, distributed, fault-tolerant solution within approximately five years of their 2023 architecture announcement. In January 2026, the company raised an additional $130 million USD in a round including RBC and TELUS, bringing total funding to $271 million and reflecting growing interest from the financial and telecommunications sectors in distributed quantum networking.

“We believe that we have correctly identified the silicon T centre as the missing component needed to finally unlock the first credible path to impactful commercial quantum computing.” — Dr. Stephanie Simmons, Photonic Inc. Founder

The Investment Landscape Transforms

Across the sector, the capital environment of 2024 and 2025 looked very different from the SPAC hangover of 2022-2023. The publicly traded quantum companies — Rigetti, IonQ, D-Wave, and others — experienced significant stock price recoveries, driven by a combination of technical milestones and renewed institutional interest in deep technology. D-Wave’s stock rose more than 3,700% over twelve months to late 2024. Rigetti and IonQ saw comparable recoveries.

Institutional investors were moving into the sector with more deliberate structures. NVIDIA announced strategic investments in Quantinuum, PsiQuantum, and QuEra in a concentrated period. JPMorgan Chase committed to investing up to $10 billion across strategic technology sectors, specifically naming quantum computing. National governments accelerated their commitments: Japan announced $7.4 billion in quantum investments in 2025; China’s national quantum programs were estimated in the tens of billions. DARPA’s Quantum Benchmarking Initiative funded eleven companies at up to $15 million each to pursue utility-scale demonstrations by 2033.

The sector had shifted from ‘is this real?’ to ‘which approaches will matter, and when?’ — a more productive, if still uncertain, question.

✦ ✦ ✦

Epilogue: The Road Ahead

By early 2026, quantum computing occupies a genuinely interesting position. It is no longer purely speculative — working quantum processors are available on the cloud, commercial contracts are being signed, and results like HSBC’s 34% improvement in bond predictions suggest that value is beginning to emerge at the edges. But the transformative applications that motivated the field’s founding — cracking encryption, simulating large molecules for drug discovery, solving global logistics problems — remain years away.

The fundamental technical challenge of error correction has not been solved at the scale required for those applications. Google’s Willow demonstrated below-threshold error correction; demonstrating it across a system large enough to run the algorithms that matter is a different order of difficulty. The path from today’s systems to the millions of physical qubits needed for fault-tolerant computation runs through engineering problems of formidable complexity.

What distinguishes 2026 from 2016 is that multiple credible paths now exist. Superconducting qubits (IBM, Google, Rigetti) have demonstrated rapid scaling and cloud deployment. Trapped ions (IonQ, Quantinuum) have demonstrated record gate fidelities and commercial revenue. Neutral atoms (QuEra, Atom Computing) have shown logical qubit error correction and surprising scalability. Photonics (PsiQuantum, Photonic Inc.) are pursuing manufacturing leverage through silicon foundries and fiber networking. And Microsoft continues its long-duration bet on topological approaches. These are not equivalent bets — they have different timelines, different risk profiles, and different target applications — but they are all being funded and built seriously.

It is also worth noting that the people drawn to this field have consistently underestimated how hard the engineering would be, while the technology has consistently delivered results that would have seemed implausible twenty years earlier. Thirty years from D-Wave’s founding, fifteen years from Google’s cloud quantum experiments, and five years from the SPAC listings that briefly made the sector look like a meme stock phenomenon, quantum computing has developed into one of the most intensely contested technology races in history.

The outcome is not certain. The timeline is not fixed. But the trajectory is real.

✦ ✦ ✦

Key Players: A Field Guide

The organizations shaping the quantum computing landscape, by architectural approach.

Superconducting Qubits

IBM Quantum The longest-running commercial quantum computing program. Superconducting transmon qubits delivered on a publicly committed multi-year roadmap, from 5 qubits on the cloud in 2016 to 1,121 on Condor in 2023. IBM Quantum Network enterprise partnerships, cloud access across 20+ systems, and a stated target of 100,000-qubit quantum-centric supercomputers by 2033.

Google Quantum AI Research-led superconducting qubit program that produced the Sycamore quantum supremacy experiment (2019) and the Willow below-threshold error correction demonstration (2024). Focused on advancing fundamental hardware science and the theoretical foundations of fault-tolerant quantum computation.

Rigetti Computing Vertically integrated superconducting qubit company operating its own quantum chip fabrication facility. Publicly listed via SPAC in 2022. Focused on hybrid quantum-classical computing and narrow quantum advantage on practical optimization problems.

Trapped Ions

IonQ Trapped-ion quantum computing company founded from the University of Maryland. First quantum hardware company to list on public markets. Ytterbium ion qubits with high gate fidelity; accessible via AWS, Google Cloud, and Azure. Growing enterprise customer base across government and commercial sectors.

Quantinuum Full-stack quantum computing company formed from the merger of Honeywell Quantum Solutions and Cambridge Quantum Computing. H-series trapped-ion processors with industry-leading two-qubit gate fidelity. Revenue-generating enterprise business in quantum chemistry, cybersecurity, and finance. The Helios system (2025) targets commercial deployment.

Neutral Atoms

QuEra Computing Neutral atom quantum computing company spun out of Harvard University, backed by Google and SoftBank. Demonstrated 48 logical qubits with Harvard researchers in 2023 — one of the first experimental demonstrations of error-corrected logical qubits outperforming physical qubits. 256-qubit Aquila system commercially available.

Atom Computing Neutral atom startup that crossed the 1,000-qubit threshold in October 2023 with 1,180 ytterbium qubits — the first gate-based system to do so. Noted for 40-second qubit coherence times and compact architecture. Subsequently partnered with Microsoft on next-generation error-corrected neutral atom systems.

Photonic Approaches

PsiQuantum Photonic quantum computing company founded in 2016 by Bristol and Imperial College London physicists. Strategy centred on manufacturing photonic quantum chips at GlobalFoundries’ semiconductor foundry and building directly toward million-qubit fault-tolerant systems. Over $2.3 billion raised, including $940 million in Australian government funding for a utility-scale facility in Brisbane.

Photonic Inc. Vancouver-based quantum computing company founded in 2016 by Dr. Stephanie Simmons. Builds on optically linked silicon spin qubits based on T-centre defects — a platform that provides both qubit computation and telecom-wavelength photonic networking in a single silicon architecture. Entanglement First distributed architecture designed to scale across existing fiber infrastructure. $271 million raised, with investors including Microsoft, RBC, and TELUS. Strategic Microsoft partnership for Azure integration. DARPA Quantum Benchmarking Initiative participant.

Quantum Annealing

D-Wave Quantum The field’s longest-standing commercial company, founded in 1999. Specializes in quantum annealing — a purpose-built approach for optimization problems rather than universal gate-based computation. First to sell quantum systems commercially (Lockheed Martin, Google/NASA). 5,000-qubit Advantage2 system available on the Leap cloud platform. Publicly listed on NYSE.

Topological Qubits

Microsoft Quantum Long-duration research program pursuing topological qubits based on Majorana zero modes — a theoretically compelling approach that would store quantum information in topologically protected states, reducing error-correction overhead. Unveiled Majorana 1 processor in February 2025. Scientific validation of topological qubit claims is ongoing. Simultaneously investing in partnerships with Photonic Inc. and Atom Computing for near-term quantum hardware development.

✦ ✦ ✦

Written March 2026. Based on publicly available research, company announcements, and peer-reviewed literature.