Quobly, a Grenoble-based silicon quantum computing company, has raised €115 million in a Series A round led by Bpifrance, SEALSQ and STMicroelectronics. The financing, one of Europe’s largest quantum rounds to date, will industrialise the company’s silicon qubits and put a first commercial system, Alloy Pioneer, on the cloud in 2026.
The deal is a bet with a clear thesis behind it. Quobly’s backers are wagering that the same 300 mm wafers that produce ordinary processors can also produce useful quantum chips, sidestepping the exotic hardware that rivals depend on. That buys manufacturing credibility most quantum startups can only describe in slides. It does not close the gap on raw qubit count, where silicon still trails.
What the €115 Million Pays For
The round was led by Bpifrance, France’s public investment bank, alongside SEALSQ and the chipmaker STMicroelectronics. The European Innovation Council (EIC), venture firm Blast, Air Liquide Venture Capital (ALIAD) and existing investor Innovacom also took part. Earlier shareholders, including CEA-Leti’s parent body the CEA, the research agency CNRS, Quantonation and Supernova Invest, remain on the cap table. It follows a roughly €19 million seed phase that ran from 2023.
The money is pointed at a single goal: moving from working prototype to a product customers can rent. Quobly says the proceeds will fund four things.
- Platform performance and scalability, raising the qubit fidelity and count the system can sustain.
- Industrialising the silicon quantum processors so they come off a fab line with consistent yield.
- Deploying the first Alloy systems in both cloud and high-performance computing (HPC) settings.
- Building out the control electronics and software stack through what the company calls system-level co-design.
One wrinkle worth flagging on the numbers. The official Quobly announcement puts the round at €115 million, while SEALSQ headlined its own participation around a €130 million figure tied to its SEALQuantum.com fund. Quobly’s primary release is the cleaner reference, so €115 million is the figure that anchors this story.
Why Quobly Builds Qubits in a Chip Fab
Most of the money in quantum computing has gone into machines that look nothing like a normal computer. Quobly’s pitch is the opposite. It wants quantum chips that roll out of the same semiconductor lines that already make billions of transistors a day, on the same 300 mm silicon wafers.
From a Transistor to a Qubit
The science behind that claim has a date attached. In 2016, a team at CEA-Leti, the French government’s microelectronics research institute, showed it could turn an industrial transistor into a working qubit by trapping a single electron and using its spin to hold quantum information. Maud Vinet, who ran CEA-Leti’s quantum program, built that work with physicists Tristan Meunier and Silvano De Franceschi, and the trio later won an ERC Synergy Grant to chase large-scale silicon quantum machines. Quobly, founded in 2022, is the company that spun out of it.
The hardware rests on fully-depleted silicon-on-insulator (FD-SOI, a low-power chip substrate technology), patterned onto 300 mm wafers using complementary metal-oxide-semiconductor (CMOS) processing, the standard recipe behind everyday chips. Quobly’s own silicon qubit technology overview frames the appeal in industrial terms: dense integration, repeatable yield and parts that behave the same from one wafer to the next.
The Scaling Wall Superconductors Hit
The incumbents have a different problem. Superconducting machines from the likes of IBM and Google have logged the headline qubit counts, but each qubit needs its own bulky control wiring inside a dilution refrigerator, and packing millions of them onto one system gets ugly fast. Silicon spin qubits are roughly the size of a transistor, so in principle they pack far tighter. Researchers recently reported industry-compatible silicon spin-qubit cells above 99% fidelity, the kind of error rate a fault-tolerant machine needs. The hard part has never been the physics of one qubit; it is making millions of them cheaply and identically. That is a fab problem, and fabs are what Europe has.
The Industrial Names Behind the Wafers
The investor list reads less like a venture syndicate and more like a supply chain. That is the point. A silicon quantum chip needs a substrate maker, a fab, a cryogenics partner and a materials supplier, and Quobly has lined up one of each, several of whom also wrote cheques.
- STMicroelectronics provides the 300 mm manufacturing line and joined the round as a lead investor, giving Quobly access to a production fab rather than a lab cleanroom.
- Soitec supplies the specialised substrates; the company’s enriched 28-silicon FD-SOI wafers are already cycling through the STMicroelectronics 300 mm fab.
- Air Liquide, through ALIAD, brings the cryogenics needed to cool the chips toward absolute zero.
- Orano contributes materials engineering, including the isotopically purified silicon that keeps qubits quiet.
- SEALSQ adds secure, post-quantum chips, folding Quobly’s processor into a wider sovereign hardware stack.
It is the kind of roster that turns a research result into something a factory can repeat. Each partner owns a piece of the process, and several now own a piece of the company.
Silicon Still Has a Qubit Gap to Close
Here is the part the funding headline glosses over. On raw qubit count, silicon spin is behind. Intel’s most advanced research chip, Tunnel Falls, carries 12 physical qubits and was shared with the academic research community rather than sold as a product, with no logical qubit yet demonstrated. Superconducting rivals count theirs in the hundreds. So the silicon camp is selling a trajectory, not a current scoreboard, and Quobly is one of four serious players making that case.
| Company | Base | Status | Recent milestone |
|---|---|---|---|
| Quobly | France | FD-SOI on 300 mm wafers; cloud system due 2026 | €115M Series A |
| Intel | United States | 12-qubit Tunnel Falls research chip | Chip shared with universities |
| Diraq | Australia | Targets a commercial machine by 2029, a million qubits by 2031 | Up to $38M U.S. CHIPS letter of intent |
| Quantum Motion | United Kingdom | Full-stack silicon CMOS system deployed at the UK national centre | $160M raise |
Two of those rivals, Diraq and Quantum Motion, advanced to Stage B of the benchmarking program run by the U.S. Defense Advanced Research Projects Agency (DARPA), a shortlist of 11 companies judged credible for a utility-scale machine by the 2030s. Quobly’s answer is to compete on the manufacturing side, where its STMicroelectronics tie gives it a working fab most peers lack.
Alloy Pioneer and the Road to the Cloud
The first product carries a name: Alloy Pioneer, the opening system in Quobly’s Alloy line, aimed at early adopters in HPC and research. Access comes through Alloy Forge, the company’s application development platform, and the rollout follows a clear sequence.
- 2016: a CEA-Leti team turns an industrial transistor into a qubit.
- 2022: Quobly is founded in Grenoble.
- 2023: a roughly €19 million seed phase funds the qubit and architecture work.
- 2026: Alloy Pioneer becomes available over the cloud via Alloy Forge.
- 2027: the system moves into HPC and data-centre infrastructure.
That last step matters more than it sounds. Quobly is designing its machines to slot into existing HPC environments rather than sit in a separate room as a science experiment, which is how it hopes to reach real industrial workloads.
A Test for European Quantum Sovereignty
The deal also carries a flag. With the EIC and Bpifrance writing cheques and STMicroelectronics, Soitec and SEALSQ supplying the hardware, this is a continental attempt to keep a strategic technology, and its fabrication, inside Europe. SEALSQ has described its stake as part of a sovereign stack running from secure chips up to quantum processors. Vinet ties the raise directly to that shift from lab to line.
With this Series A, we are accelerating the deployment of our first commercial systems and building a quantum computing platform designed to integrate into existing computing infrastructures. Our objective is to make quantum computing deployable, scalable and usable within real industrial environments.
That was Maud Vinet, CEO and co-founder of Quobly. The EIC has tracked the company since its early funding, and its profile of the silicon quantum venture sets the same goal of European-built, scalable machines. The first real test arrives when Alloy Pioneer goes live on the cloud in 2026, and HPC users get to benchmark French silicon against everything else on the market.
Frequently Asked Questions
What is Quobly?
Quobly is a French quantum computing company founded in 2022 and based in Grenoble. It spun out of CEA-Leti, the French government microelectronics institute, and builds quantum processors from silicon qubits manufactured on standard 300 mm semiconductor wafers.
How does silicon quantum computing differ from superconducting machines?
Silicon spin qubits store quantum information in the spin of single electrons trapped in transistor-like structures, so they are tiny and can be made in existing chip fabs. Superconducting qubits, used by IBM and Google, are larger circuits that need extensive cryogenic wiring, which makes packing millions of them harder.
When will Quobly’s Alloy Pioneer be available?
Alloy Pioneer is expected to be accessible through the cloud in 2026 via Quobly’s Alloy Forge platform, aimed at early adopters in high-performance computing and research, before deployment into HPC infrastructure in 2027.
Who led Quobly’s €115 million Series A?
The round was led by Bpifrance, SEALSQ and STMicroelectronics, with participation from the European Innovation Council, Blast, Air Liquide Venture Capital and existing investor Innovacom. The CEA, CNRS, Quantonation and Supernova Invest remain shareholders.
What is FD-SOI technology?
FD-SOI, or fully-depleted silicon-on-insulator, is a low-power chip substrate technology already used in commercial semiconductors. Quobly patterns its qubits onto FD-SOI wafers using standard CMOS processing, which is how it aims to get repeatable yield and dense integration.
