May 21, 2026 – Every year, governments around the world announce quantum funding packages, and every year the coverage follows the same script: big number, stock tickers, vague language about “leadership.” Yesterday’s $2 billion CHIPS Act announcement from the U.S. Commerce Department looks like more of the same — until you read what the money is actually for.
The $2 billion is an industrial policy intervention aimed at building the physical manufacturing infrastructure that quantum computers will need. Two-thirds goes to foundries: $1 billion to IBM to stand up Anderon, America’s first dedicated quantum wafer foundry, and $375 million to GlobalFoundries to launch a quantum manufacturing division covering cryo-CMOS, advanced packaging, and superconducting interconnects. The remaining third funds seven quantum computing companies to solve specific engineering bottlenecks. Research grants fund papers. This money funds fabs and supply chains.
The bottom line: when a country starts building purpose-built quantum fabrication facilities, it is no longer asking whether the technology works. It is preparing to produce it at scale. For anyone tracking the CRQC timeline or managing a PQC migration program, that shift in posture should sharpen the sense of urgency.
On May 21, 2026, the Department of Commerce signed letters of intent with nine companies for $2.013 billion in federal incentives under the CHIPS and Science Act. The government takes a minority, non-controlling equity stake in every recipient.
Foundry investments ($1.375 billion):
Quantum computing company investments ($638 million across 7 companies):
All awards are subject to execution of definitive award documents.
The Commerce Department framed this as a portfolio approach, and the portfolio design reveals careful thinking. Two foundries provide the manufacturing substrate. Seven companies spanning five qubit modalities address what the CHIPS R&D Office calls “the most consequential, unresolved engineering problems.” The NIST announcement names specific technical targets: device reproducibility, optical complexity, error rates, cryogenic systems integration, control hardware, ultra-fast readout electronics, photonic loss, and interconnects.
Each of the seven company awards is pointed at a discrete engineering bottleneck, not at general-purpose quantum research. That specificity separates this from the dozens of government quantum programs that amount to “here’s money, go do quantum.” When the Commerce Department tells Rigetti to work on miniaturized readout electronics and next-generation cryostat architectures, or tells PsiQuantum to solve electro-optic materials and single-photon detector performance, it is prescribing manufacturing-grade engineering problems.
IBM’s Anderon is a 300 mm quantum wafer foundry designed to serve multiple quantum hardware vendors. IBM will contribute $1 billion of its own capital alongside the government’s $1 billion, plus intellectual property, manufacturing assets, and workforce. The company positions Anderon as a pure-play foundry, a standalone entity that will fabricate quantum wafers for the broader industry, not just for IBM’s internal roadmap.
If this model works, it could do for quantum what TSMC did for classical semiconductors: separate chip design from chip manufacturing so that a broader ecosystem of quantum hardware companies can scale without each building its own fab. Today, most quantum QPU companies either fabricate in-house (IBM, Google), rely on academic clean rooms, or contract with general-purpose fabs like GlobalFoundries. A dedicated quantum foundry changes the economics by spreading fabrication costs across multiple customers.
The comparison to TSMC is aspirational and premature. Quantum wafer volumes are orders of magnitude smaller than classical semiconductor volumes, and the process requirements (Josephson junctions, through-silicon vias for 3D wiring, superconducting interconnects) are specialized enough that “general-purpose quantum foundry” may be an oxymoron for years. But the structural ambition is clear, and IBM has the fabrication expertise to attempt it.
While Anderon focuses on superconducting quantum wafer fabrication, GlobalFoundries’ $375 million funds a multi-modality manufacturing platform. GF’s announcement names quantum processor units, cryogenic control and readout ICs, advanced packaging, and superconducting interconnects across superconducting, trapped ion, photonic, topological, and silicon spin modalities. Companies including Diraq, PsiQuantum, Quantinuum, Google, Microsoft, and NVIDIA are already engaged with GF’s manufacturing platform.
This is the cryo-CMOS and advanced packaging play. As I’ve covered in my CRQC Quantum Capability Framework, the path to engineering scale and manufacturability depends not just on qubit chips but on the surrounding classical infrastructure that must operate at cryogenic temperatures. GF has spent years building cryogenic CMOS expertise with quantum companies, and that positions it to manufacture the control electronics that sit inside the dilution refrigerator alongside the qubits. I’ve tracked this bottleneck closely: as qubit counts scale past a few hundred, the I/O wall between room-temperature electronics and millikelvin processors becomes the binding constraint, and cryo-CMOS is one of the few credible paths through it.
The seven quantum computing recipients cover an admirably broad modality spread, but the lineup has conspicuous gaps.
The inclusions that make sense: Quantinuum ($100M) is arguably the most technically advanced quantum company on Earth, with its Helios system demonstrating 48 logical qubits at a 2:1 physical-to-logical ratio. PsiQuantum ($100M) has the most ambitious manufacturing thesis in quantum, with its Omega chipset fabricated at GlobalFoundries. Atom Computing ($100M) holds the coherence record for neutral atoms. Infleqtion ($100M) brings neutral-atom engineering depth. Rigetti ($100M) has the most mature Quantum Open Architecture QPU offering alongside QuantWare. Diraq ($38M) is the only silicon spin company on the list, and its September 2025 Nature paper demonstrating >99% two-qubit fidelity on industrial 300 mm wafers validated the most scalable manufacturing path in quantum.
The inclusion that raises questions: D-Wave ($100M) is described as receiving funding for “annealing and gate-model superconducting quantum computing systems.” D-Wave has real commercial traction with its annealing platform; it is the only quantum company with meaningful recurring revenue from actual quantum hardware customers. But its gate-model program is nascent, and the company’s stated targets of a 100,000-qubit annealing system and a 10,000-qubit gate-model system are both ambitious roadmap items. The gate-model target in particular, “dozens of logical qubits,” puts D-Wave years behind Quantinuum and IBM in the fault-tolerant race. Whether $100 million accelerates that trajectory or sustains a long-shot bet is an open question.
The notable absence: IonQ, the highest-revenue publicly traded pure-play quantum company, is not on the list. IonQ reported $64.7 million in Q1 2026 revenue (a 755% year-over-year increase), recently achieved 99.99% two-qubit gate fidelity (a world record), and is in the process of acquiring SkyWater Technology for $1.8 billion, which would give it its own U.S.-based foundry. The market didn’t seem bothered: IonQ shares rose 11% on the day, apparently pricing in the logic that a company sitting on $3.1 billion in cash and about to own its own fab doesn’t need a $100 million government grant. Still, IonQ’s absence from a government portfolio designed to cover multiple modalities, with no trapped-ion company other than Quantinuum, is conspicuous.
Also absent: QuEra, which demonstrated 96 logical qubits and recently achieved 2:1 physical-to-logical qubit ratio, and has one of the most credible neutral-atom roadmaps. The entire quantum networking and communication layer is missing too (no Qubitekk, no ID Quantique, no Aliro Quantum), though the CHIPS Act’s scope arguably doesn’t cover networking.
The Commerce Department will take a minority, non-controlling equity stake in each of the nine recipients. This follows the precedent set by the Trump administration’s August 2025 conversion of CHIPS Act grants to Intel into a 9.9% equity stake, and the Pentagon’s 15% stake in rare-earth producer MP Materials.
For the quantum companies, the equity mechanism has different implications depending on the recipient. IBM and GlobalFoundries are large-cap public companies where a minority government stake is a rounding error (the government gets approximately 1% of GlobalFoundries). For the smaller companies (D-Wave, Rigetti, Infleqtion, Diraq, Atom Computing), $100 million in shares issued to the government represents meaningful dilution. D-Wave’s announcement specifies it will issue $100 million in common stock to the Commerce Department.
The government’s Intel stake is now reportedly worth approximately $36 billion, up from $8.9 billion at acquisition, a return that the Council on Foreign Relations has called one of the most profitable government investments in American industrial history. Whether quantum equity stakes generate similar returns depends entirely on whether these companies build what they say they’re going to build. That’s a longer bet than Intel, which was already a $200 billion company when the government bought in.
Zoom out from Washington, and a pattern comes into focus: everyone is converging on manufacturing as the bottleneck, simultaneously, from different directions.
Two weeks before this CHIPS announcement, QuantWare raised $178 million in the largest private round ever by a dedicated quantum processor company, backed by Intel Capital and In-Q-Tel. The money funds KiloFab, a dedicated quantum fabrication facility in Delft that will increase production capacity by 20×. In March 2026, the UK committed £2 billion ($2.67 billion) under the ProQure program, structured around procurement and manufacturing: companies build prototype systems, the government evaluates them, and the best designs scale into national computing infrastructure. Japan committed $7.4 billion, with its 18 Bluefors KIDE cryostats at the AIST G-QuAT center in Tsukuba representing the single largest cryogenic manufacturing procurement in quantum history. And China’s Origin Quantum has been building its own domestic supply chain, dilution refrigerators, measurement-control systems, programming frameworks, and a downloadable operating system, precisely because it cannot access Western manufacturing infrastructure under export controls.
Three years ago, quantum funding stories were about algorithms, cloud access, and qubit counts. Today they’re about fabs, foundries, and wafer processes. That shift tells you where the field thinks the real constraint lives.
This CHIPS package also marks a structural shift in how the U.S. funds quantum technology, and the contrast with China is instructive.
As I’ve covered extensively in my China’s Quantum Ambition series, China’s quantum program has been overwhelmingly state-driven from the start. Public funds dwarf private capital. A 2022 RAND Corporation report found that China’s private sector quantum funding was only 3% of U.S. corporate and venture capital. The Chinese Academy of Sciences controls the research agenda, USTC trains the talent, and companies like Origin Quantum and QuantumCTek are state-backed spin-outs. This model produces coordination, speed, and vertical integration at the cost of commercial competition and ecosystem diversity.
The U.S. model, until now, has been the opposite: private capital and corporate R&D do the heavy lifting, with government providing research grants, export controls, and strategic direction. American quantum companies raised $2.7 billion of the $4.9 billion in global quantum VC in 2025. IBM, Google, Microsoft, IonQ, Rigetti, and Quantinuum built their platforms with shareholder and VC money, not government grants. The strength is commercial dynamism and competitive pressure. The weakness is that no single VC-backed startup can afford to build a quantum wafer foundry. The capital requirements and uncertain return timelines don’t fit the venture model.
The CHIPS quantum package creates a hybrid. The government provides manufacturing-scale capital ($1.375 billion in foundry investment) that the private market wouldn’t fund, while the venture-backed companies identify and solve the specific engineering problems. The equity stake mechanism aligns government and company incentives: the taxpayer participates in the upside if the technology succeeds. The result is targeted public capital addressing a specific market gap (manufacturing infrastructure) while the private sector retains its role in technology development and commercialization.
Whether this hybrid outperforms either pure model remains to be seen. The UK’s ProQure program, QuantWare’s KiloFab, and Japan’s AIST G-QuAT procurement all follow a similar pattern: public money building the manufacturing base, private companies building the technology on top of it. Five countries converging independently on the same funding architecture, within the span of a year, suggests that the industry has collectively reached the same conclusion about where the bottleneck is.
In raw dollar terms, $2 billion is significant but not dominant. The QED-C’s State of the Global Quantum Industry 2026 report estimates $56.7 billion in cumulative global public quantum funding, with $12.7 billion added in the past year alone. China’s quantum investment is opaque but estimated at $15 billion or more. Japan committed $7.4 billion. The UK’s total quantum program now exceeds £4.5 billion ($6 billion). France has committed €1.8 billion. South Korea announced ₩3 trillion (~$2.2 billion).
But the comparison is misleading if you only look at the topline number. Most of that global public funding went to research institutions, university programs, and early-stage grants. The CHIPS quantum package is almost entirely manufacturing-oriented. The relevant comparison is not “$2 billion versus $56.7 billion in total quantum funding” — it is “$1.375 billion in quantum foundry investment versus near-zero in dedicated quantum manufacturing infrastructure spending by any other single government in a single tranche.” On that metric, this is the largest bet any Western government has placed on quantum manufacturing capacity.
From a Q-Day and CRQC timeline perspective, this investment is more relevant than it might appear at first glance. The path to a cryptographically relevant quantum computer is constrained less by algorithmic breakthroughs than by engineering scale and manufacturability: the ability to reliably fabricate thousands of high-quality qubits and integrate them with classical control systems. That’s exactly what foundry investment addresses.
A dedicated quantum foundry producing wafers at 300 mm scale, combined with cryo-CMOS manufacturing at GlobalFoundries, accelerates the timeline on which physical qubit counts can scale into the thousands and tens of thousands. The physics hasn’t changed: two-qubit gate fidelities still need to improve, quantum error correction overhead still needs to shrink, and decoder performance still needs to reach real-time speeds. But the announcement removes one of the bottlenecks, the lack of fabrication infrastructure purpose-built for quantum devices.
For CISOs and security leaders tracking the quantum threat, this announcement should reinforce the urgency I’ve been arguing: the ecosystem is accelerating. Government money flowing into manufacturing infrastructure is a later-stage signal than government money flowing into research. Countries build fabs when they’ve decided a technology is going to work and want to own its production. The regulatory and ecosystem deadlines driving PQC migration don’t wait for these fabs to come online.
These are letters of intent, not final awards. The Commerce Department is clear: definitive award documents still need to be executed. The Intel CHIPS deal went through multiple iterations before finalizing. The same could happen here, and the amounts could change.
Beyond the paperwork, several questions will determine whether this $2 billion produces lasting strategic value:
Will Anderon actually operate as a multi-customer foundry, or will it default to serving IBM’s internal roadmap? The answer determines whether this is an industry investment or a company subsidy.
Can GlobalFoundries’ Quantum Technology Solutions demonstrate cryo-CMOS manufacturing at the volumes and yields that quantum scaling requires? Cryogenic CMOS operates under different physical constraints than room-temperature semiconductors, and the technical challenge is real.
Does the modality diversity in the portfolio translate to hedged bets, or does it spread capital too thin? $100 million is meaningful for a startup, but it won’t build a fault-tolerant quantum computer. It can, however, solve a specific engineering bottleneck, and the NIST announcement suggests that’s exactly how the CHIPS R&D Office is thinking about it.
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。