Quantum's commercial moment arrives at VivaTech — but the hardware still cannot keep up

At VivaTech 2026 in Paris on 18 June 2026, the pitch from the quantum computing industry was unusually direct: stop treating this as a science project. The technology, exhibitors and panellists argued, is ready to do useful work in drug discovery, optimisation, finance and materials science — and the bottleneck is no longer the physics, but the procurement decisions of large customers who have not yet learned to buy it.

That framing is convenient, and partially true. It is also a useful place to start asking harder questions — about how many machines actually exist, who owns them, and whether the supply chain that would have to scale production is anywhere close to ready.

The promise, restated for the boardroom

The headline case for quantum has not changed in years. Classical computers struggle with certain classes of problem — simulating molecular interactions, optimising logistics networks, factoring large integers — that quantum machines, in principle, handle natively. A useful quantum computer with a few thousand error-corrected logical qubits would, on paper, outperform the largest classical supercomputers at specific tasks.

At VivaTech, the discussion reportedly shifted from the underlying physics to commercial pilots. The implication was that the era of laboratory demonstrations is ending, and a market is beginning. For a technology that has been "ten years away" for roughly three decades, that is a significant rhetorical move.

The hardware ledger

The harder question is supply. Useful, error-corrected quantum machines remain vanishingly rare. The most credible commercial systems in operation today are still largely cryogenically cooled superconducting devices with tens to low hundreds of physical qubits, prone to noise, and lacking the error-correction overhead that would let them run algorithms of commercial length. Trapped-ion and neutral-atom platforms have made progress on qubit quality, but have not solved the scaling problem either.

Three constraints dominate. First, fabrication: producing the control electronics, dilution refrigerators and photonic interconnects at the volumes a real industry would require is not yet a solved manufacturing problem. Second, talent: the pipeline of engineers who can build and maintain these systems is thin, and concentrated in a handful of national labs and corporate research groups. Third, integration: most useful workloads will require hybrid classical-quantum architectures, and the software stack to manage that hybrid is still in early days.

None of this is a secret inside the industry. The interesting question is why the commercial pitch has nevertheless hardened in 2026.

The capital question

The honest answer is that capital is patient, and government is patient, but patience is not infinite. Public funding for quantum research has been substantial — in the United States, the European Union, the United Kingdom, China and France — and the question now being asked, in budget meetings as much as in conference halls, is when that spending starts to produce measurable economic return.

That pressure is producing two effects. It is accelerating the search for near-term commercial applications — small, noisy machines doing useful work in chemistry or optimisation without requiring full error correction. And it is forcing a degree of honesty about timelines that the field has historically avoided. A useful, fault-tolerant general-purpose quantum computer remains, on the evidence available, a multi-year engineering programme rather than a product on a price list.

There is a third effect, less discussed: a quiet consolidation. The number of independent quantum hardware start-ups is shrinking, and the survivors are those with credible paths to fabrication scale, not just to a good paper.

The industrial policy frame

Quantum is no longer just a research story; it is an industrial policy story. The United States has restricted the export of certain quantum-related components to China. The European Union has funded quantum pilot lines as part of its broader semiconductor and chips strategy. France is positioning itself as a continental hub, and VivaTech is part of that marketing. China has parallel programmes in superconducting and photonic quantum hardware, and is, by several measures, at or near the frontier in specific sub-fields.

This matters because the commercial case for quantum is being made inside a geopolitical frame. Customers are being asked to commit to long procurement timelines, and to integrate quantum components into workflows that may not deliver economic returns for years. The customers who can absorb that uncertainty are largely states, defence agencies, and a small number of very large financial and pharmaceutical firms. The market, in other words, is not yet a market in the conventional sense.

What remains uncertain

The VivaTech pitch assumes that the gap between today's noisy intermediate-scale machines and tomorrow's fault-tolerant systems can be closed without a discontinuous event — a fabrication breakthrough, a materials insight, or a new error-correction scheme that collapses the required qubit count. None of those can be ruled out. None can be scheduled.

The sources available from this year's conference do not specify commercial shipment volumes, do not name a price point, and do not name a customer outside the early-access cohort that has, by industry practice, been receiving pro bono access in exchange for case studies. That is itself a tell. A market that exists tends to publish prices.

What is genuinely new in 2026 is the seriousness with which the question is being asked. The conference, and the broader European quantum push around it, is a signal that policymakers have decided the technology is strategically important even if its commercial returns remain uncertain. That is a defensible position. It is also one that the industry, given the opportunity, will use to keep the public money flowing for as long as the framing holds.