May 22, 2026 – One day after the U.S. committed $2 billion under the CHIPS Act to build quantum manufacturing infrastructure, French President Emmanuel Macron announced an additional €1 billion for France’s quantum strategy plus €550 million for the microelectronics sector. “The speed of our competitors requires that we shift into a higher gear,” Macron said during a visit to a supercomputing center south of Paris. The additional commitment brings France’s total public quantum investment to €3.3 billion, up from the €2.3 billion already committed since the original €1.8 billion national strategy launched in 2021.
The same day, NVIDIA’s venture capital arm NVentures invested in Alice & Bob, the Paris- and Boston-based company building fault-tolerant quantum computers on a proprietary cat-qubit architecture. The investment extends Alice & Bob’s €100 million Series B, which closed in January 2025 with backing from Future French Champions, AXA Venture Partners, and Bpifrance (France’s public investment bank). Financial terms were not disclosed.
The bottom line: the US commits $2 billion on Wednesday, France responds with €1 billion on Thursday, and NVIDIA quietly buys into a French quantum champion the same afternoon. The quantum manufacturing race is now moving on a 24-hour news cycle.
Alice & Bob specializes in cat qubits, a superconducting qubit design where bit-flip errors are suppressed exponentially at the hardware level. The company claims this built-in error protection could reduce the physical qubit overhead for fault-tolerant computing by up to 200× compared to conventional transmon architectures. Using LDPC codes on cat qubits, Alice & Bob projects that 100 high-fidelity logical qubits could run on as few as 1,500 physical cat qubits. The company targets a useful fault-tolerant quantum computer by 2030.
NVentures has spent the past nine months writing cheques across every major hardware approach. The fund participated in Quantinuum’s $600 million round (trapped ion, $10 billion valuation), invested in QuEra (neutral atom), joined PsiQuantum’s $1 billion Series E (photonic), and now adds Alice & Bob (superconducting cat qubit). NVIDIA’s strategy is clear: rather than build its own QPU, it positions CUDA-Q and NVQLink as the classical compute and software substrate that every quantum modality will need to run alongside. Alice & Bob has already integrated its architecture with CUDA-Q, cuQuantum, and NVQLink.
Alice & Bob sits inside PROQCIMA, France’s flagship quantum hardware program. PROQCIMA, launched in March 2024 under the Direction Générale de l’Armement (DGA, France’s defense procurement agency), targets two prototypes of universal fault-tolerant quantum computers with 128 logical qubits by 2030 and industrialized 2,048-logical-qubit systems by 2035. Five French companies share up to €500 million in Phase 1 funding: Alice & Bob (cat qubits), Pasqal (neutral atoms), Quandela (photonics), Quobly (silicon spin), and C12 Quantum Electronics (carbon nanotubes).
On the manufacturing front, Alice & Bob is constructing a €50 million advanced quantum lab in Paris with a nanofabrication cleanroom for prototyping cat-qubit chips. In October 2025, the French defense innovation agency AID funded the delivery of PLASSYS-BESTEK’s SQUID-6 UHV, a fully automated ultra-high vacuum deposition system for quantum chip fabrication, to this facility under the ULTRACAT project.
I wrote yesterday that five countries had converged independently on the same conclusion within a single year: the quantum bottleneck is manufacturing. Macron’s response within 24 hours confirms the pattern and accelerates it. France isn’t just continuing an existing strategy. It is scaling up in direct response to the US move.
The PROQCIMA program is structured as a competition to build quantum computers, with hardware deliverables and industrial milestones, not as a research grant for publishing papers. The DGA leads it. The program name itself is a reference to ULTRA, the British World War II codebreaking program. When a country’s defense procurement agency runs the quantum computer program and names it after history’s most famous cryptanalysis effort, the strategic intent is not subtle.
Alice & Bob’s €50 million Paris lab with its own nanofabrication cleanroom follows the same logic as IBM’s Anderon foundry in Albany and QuantWare’s KiloFab in Delft: companies that want to scale quantum hardware are building dedicated fabrication facilities. Each investment is different in scale (Anderon’s $2 billion dwarfs Alice & Bob’s €50 million), but the direction is identical. The era of quantum companies prototyping chips in borrowed university clean rooms is ending.
NVentures’ portfolio now spans trapped ion (Quantinuum), neutral atom (QuEra), photonic (PsiQuantum), and superconducting cat qubit (Alice & Bob), on top of NVIDIA’s direct partnerships with superconducting transmon vendors through NVQLink. Timothy Costa, NVIDIA’s VP and GM of Quantum, framed the investment around a shared vision for “accelerated quantum supercomputing” that connects quantum processors to classical GPU infrastructure.
The logic is straightforward. NVIDIA doesn’t know which qubit modality will reach fault tolerance first, and it doesn’t need to. If CUDA-Q and NVQLink become the standard interface between QPUs and classical HPC (as NVQLink’s 17 quantum hardware partners and 9 U.S. national laboratories suggest they might), NVIDIA collects a toll on every quantum computation regardless of which qubit technology wins. The venture investments ensure that the leading company in each modality is already integrated with NVIDIA’s software before the fault-tolerance race is decided.
Alice & Bob’s 200× overhead reduction claim deserves scrutiny. The company’s January 2024 arXiv paper on LDPC codes for cat qubits projects 100 logical qubits from 1,500 physical cat qubits. For comparison, conventional surface-code transmon architectures require roughly 1,000 physical qubits per logical qubit at useful error rates (distance ~20). If the cat-qubit numbers hold, the path to fault tolerance shortens considerably.
The catch is that cat qubits suppress bit-flip errors at the cost of increased phase-flip errors, and the full error correction architecture must handle both. Alice & Bob’s approach uses a repetition code for the remaining phase-flip errors (since bit flips are already suppressed), which they argue is simpler and more hardware-efficient than the full 2D surface code. The theoretical argument is sound, and the company demonstrated a 7-minute bit-flip lifetime in 2024. But no one has yet demonstrated a fault-tolerant logical operation on cat qubits. The 200× figure is a projection from theoretical architecture papers, not a measured result on hardware.
Amazon’s AWS has adopted the cat-qubit approach for its own quantum hardware program, which provides independent validation of the theoretical premises. Whether the engineering delivers on the theory remains the open question.
France’s quantum strategy has a coherence that other national programs lack. €3.3 billion in total public commitment. Five hardware companies spanning five distinct qubit modalities (cat qubit, neutral atom, photonic, silicon spin, carbon nanotube), all funded through a single competitive program with defense oversight, industrial milestones, and a 10-year timeline to 2,048 logical qubits. Bpifrance sits on every cap table. The DGA sets the hardware deliverables. STMicroelectronics and CEA provide the semiconductor manufacturing ecosystem. And Macron personally announced the latest €1 billion tranche the day after the US moved, framing it as a sovereignty issue: “Technological dependencies will more and more become industrial and strategic dependencies.”
The risk is the one Bpifrance CEO Nicolas Dufourcq flagged publicly in September 2025: European private capital is underinvesting in quantum relative to the US and China. France’s quantum companies depend heavily on sovereign capital (Bpifrance, DGA, FFC, the Qatar-backed Future French Champions fund). The NVentures investment in Alice & Bob is notable precisely because it is non-French private capital validating a French quantum company. If European VC doesn’t follow, France’s quantum champions will hit a funding ceiling that government money alone cannot breach.
The US CHIPS Act announcement on May 21 and Macron’s €1 billion on May 22 are not coincidental. Macron framed the investment as a direct competitive response. Quantum industrial policy has entered a feedback loop: one country moves, another responds within a day, and the pace of commitment ratchets upward. The NVentures investment in Alice & Bob landing the same afternoon adds a private-capital signal on top of the sovereign one.
My company - Applied Quantum - helps governments, enterprises, and investors prepare for both the upside and the risk of quantum technologies. We deliver concise board and investor briefings; demystify quantum computing, sensing, and communications; craft national and corporate strategies to capture advantage; and turn plans into delivery. We help you mitigate the quantum risk by executing crypto‑inventory, crypto‑agility implementation, PQC migration, and broader defenses against the quantum threat. We run vendor due diligence, proof‑of‑value pilots, standards and policy alignment, workforce training, and procurement support, then oversee implementation across your organization. Contact me if you want help.
此内容由惯性聚合(RSS阅读器)自动聚合整理，仅供阅读参考。 原文来自 — 版权归原作者所有。