Most quantum companies are climbing the ladder one rung at a time, building bigger noisy machines and hoping to reach reliability eventually. PsiQuantum decided to skip the ladder entirely and build, from the start, the only kind of machine it believes will matter: a million-qubit, fault-tolerant computer.
Founded in 2016 in Palo Alto by a team of physicists from the University of Bristol and Imperial College London, including Jeremy O'Brien, Terry Rudolph, Pete Shadbolt, and Mark Thompson, PsiQuantum made one of the boldest bets in the field. Rather than chase the near-term milestones that generate headlines, it set out to solve the end-state problem directly, on the conviction that anything short of a large, error-corrected machine will never deliver the transformative applications quantum computing promises. In early 2026 it brought in former AMD president Victor Peng to lead the company through its shift from research toward large-scale deployment, with O'Brien moving to executive chairman.
It is a high-risk, high-ceiling strategy, and it rests on a single distinctive choice: building quantum computers out of light.
PsiQuantum builds its machines using photons, individual particles of light, as qubits, guided through circuits etched onto silicon photonic chips. This approach is fundamentally different from the superconducting circuits and trapped ions that dominate much of the field, and it brings a set of advantages that PsiQuantum believes are decisive at scale.
Photons interact weakly with their environment, which can help with certain kinds of stability, and much of the system operates without the elaborate, near-absolute-zero refrigeration that superconducting machines require, with cryogenic cooling needed mainly for the detectors. Just as importantly, photons are the natural carriers of information over optical fiber, which means a photonic quantum computer is inherently suited to networking, linking modules together and, eventually, connecting machines across distances.
That networkability is central to PsiQuantum's vision. A million-qubit machine will not be a single monolithic chip. It will be a vast, interconnected system, and a technology in which the qubits move at the speed of light through fiber is well matched to building it. Where some approaches must invent new ways to connect their pieces, photonics offers connection as a native property.
The trade-off is that manipulating single photons reliably is extraordinarily demanding, and photonic approaches have historically faced steep engineering challenges. PsiQuantum's wager is that these challenges are solvable with the right architecture and, crucially, the right manufacturing base, and that the long-term advantages of light will win out.
It is a contrarian position, but a coherent one, and the company has pursued it with unusual single-mindedness from its earliest days.
PsiQuantum's most important strategic asset may be how its chips are made. Rather than fabricating delicate devices by hand in a laboratory, the company manufactures its silicon photonic chips in a conventional semiconductor foundry, working with GlobalFoundries at a commercial fab. This means PsiQuantum's quantum chips roll off the same kind of advanced production line that makes the world's ordinary computer chips.
The significance of this is hard to overstate. The single hardest problem in scaling quantum computing is manufacturing an enormous number of high-quality components reliably and affordably. A brilliant design that can only be built one device at a time in a research setting is a dead end on the path to a million qubits. By anchoring its technology in existing, proven semiconductor manufacturing, PsiQuantum aims to leap over that obstacle.
This fab-based approach gives the company a credible answer to the question every quantum effort must eventually face: how will you actually build this at scale? PsiQuantum's answer is that the infrastructure already exists, refined over decades of making conventional chips, and that its job is to harness it for quantum. That is a genuinely differentiated position in the field.
It also means PsiQuantum benefits from the relentless improvement of the broader semiconductor industry, rather than having to invent every piece of its manufacturing process from scratch. Standing on the shoulders of the world's chipmaking capability is a powerful place to be.
The deepest distinction of PsiQuantum's strategy is philosophical. Much of the field has spent years building noisy, intermediate-scale machines, useful for research and experimentation but not for the large, reliable computations that justify the technology. PsiQuantum deliberately skipped that phase, designing from the outset for fault tolerance, the error-corrected reliability that practical applications require.
Its architecture, built around a concept called fusion-based quantum computing, was designed with error correction and tolerance for the inevitable loss of photons baked in from the beginning. Rather than building a small machine and figuring out error correction later, PsiQuantum structured its entire approach around the requirements of a large, fault-tolerant system, then worked backward to the components needed to build it.
This avoids what some call the noisy-machine dead end, the risk of investing heavily in intermediate systems that never lead to the genuinely useful machine. PsiQuantum bet that the only destination worth reaching is the fault-tolerant one, and that the most direct route is to aim straight for it.
The approach demands patience and conviction, because it means less to show in the short term than rivals who can demonstrate steadily larger noisy machines. But if the bet pays off, PsiQuantum could arrive at genuine utility without the detours, and its single-minded focus on the end goal is itself a kind of clarity that the field respects.
It is, in essence, a bet that in quantum computing the only milestone that ultimately matters is the last one, and that everything should be organized around reaching it.
In early 2025, PsiQuantum gave the field a concrete look at its progress with the unveiling of its Omega chipset, described in a paper in the journal Nature. Omega is a manufacturable set of silicon-photonic components, the building blocks needed to assemble large-scale machines, produced in a commercial fab. It was a demonstration that the company's foundational technology works and can be made the way PsiQuantum has always promised.
The importance of Omega is that it moves the company's vision from architecture on paper toward components in hand. It shows that the pieces of a photonic quantum computer, the sources, the circuits, the detectors, can be fabricated together with the quality and consistency that scaling requires. For a company whose strategy depends entirely on manufacturability, that is a crucial proof point.
Publishing the work in a leading peer-reviewed journal also gave the result scientific credibility, signaling that PsiQuantum's claims rest on verified results rather than ambition alone. It was a meaningful step in turning a bold, long-horizon bet into a demonstrable engineering program.
What truly sets PsiQuantum apart at this stage is that it is already building the physical homes for its future machines. The company has broken ground on two flagship sites intended to house utility-scale, fault-tolerant quantum computers: one in Brisbane, Australia, backed by roughly a billion Australian dollars in government funding, and one in the Chicago area, positioned as the first utility-scale quantum site in the United States.
These are not modest installations. They are major facilities, complete with the enormous cooling infrastructure such systems require, developed through partnerships including one for what is planned to be among the largest cryogenic plants ever built for a quantum computer. Committing to construction on this scale reflects a company that believes its machine is close enough to start building the infrastructure around it.
The government backing is just as telling. National and regional governments do not commit funding of this magnitude lightly, and their willingness to anchor PsiQuantum's sites signals serious confidence in the company's approach, along with a desire to host strategically important quantum capability. These deals provide both capital and a built-in mandate, reducing the company's risk while accelerating its timeline.
By building the sites in parallel with the machines, PsiQuantum is compressing the path from a working system to a deployed one. When its technology is ready, the facilities to house it will already exist, which is a distinctive advantage in a field where most players are still focused only on the hardware itself.
A strategy as ambitious and long-horizon as PsiQuantum's requires deep pockets, and the company has assembled an exceptional roster of backers. In 2025 it raised a major funding round on the order of a billion dollars, led by some of the world's most sophisticated long-term investors, bringing its valuation into the multibillion-dollar range and making it the best-funded pure photonic quantum company by a wide margin.
This capital matters because PsiQuantum's bet only pays off at the end, when the large machine is complete, which means it must fund years of work before reaching its destination. A blue-chip investor base willing to take that long view is exactly what such a strategy demands, and the company's ability to attract it is a vote of confidence in both the team and the approach.
The arrival of an experienced semiconductor and technology leader at the helm signals a shift toward execution and scale, the right posture for a company moving from proving its science to building its machines. It is the kind of leadership a capital-intensive, manufacturing-driven effort needs as it industrializes.
Together, the funding, the government partnerships, and the manufacturing base give PsiQuantum the resources and the runway to pursue its singular goal without the constant pressure of near-term commercialization that constrains many rivals.
PsiQuantum represents the boldest version of the quantum dream: not a bigger noisy machine, but a direct assault on the fault-tolerant, million-qubit computer that could actually transform chemistry, materials, medicine, and more. Its photonic, fab-based, fault-tolerance-first approach is genuinely differentiated, and its willingness to build utility-scale sites before the machine is finished reflects extraordinary conviction.
For anyone watching where quantum computing might break through, PsiQuantum is one of the most fascinating bets in the field. It is high risk, because so much rests on reaching the end state, but the ceiling is enormous, because if it succeeds it could leap directly to the kind of machine everyone else is still climbing toward. In a field full of incrementalists, PsiQuantum is swinging for the fences, and it has the manufacturing base, the capital, and the government backing to make the swing count.
It is worth being clear about what is at stake in PsiQuantum's wager. If the company succeeds, it would not merely add another machine to the field. It would deliver the kind of large, reliable quantum computer that researchers have spoken about for decades, the machine capable of designing new medicines and materials from first principles and solving problems that no classical computer ever could. The prize is not incremental. It is the whole point of the field.
That is why PsiQuantum's approach, for all its risk, commands such attention. The company has refused to settle for impressive demonstrations along the way, insisting instead on the destination that actually matters. Combined with a manufacturing strategy rooted in real fabs, marquee government partnerships, and a deep base of patient capital, that focus makes PsiQuantum one of the most consequential bets in quantum computing. Whether or not it arrives first, it is pursuing the version of success that would change the most.
Jason Kumpf follows the quantum industry for what it means to business. He is Head of US Revenue at Razorpay, a board advisor, angel investor, and speaker. More about Jason.