Commerce signed nine LOIs worth $2.013 billion to anchor a U.S. quantum supercomputing supply chain. Two foundry models funded in parallel, federal equity in every recipient.
The U.S. Department of Commerce announced on May 21 that it has signed nine letters of intent worth $2.013 billion in CHIPS R&D incentives with two domestic quantum foundry operators and seven quantum hardware developers. Two structural features set this package apart from the high-profile CHIPS-era deals that preceded it. First, Commerce is positioning to fund two foundry architectures in parallel rather than picking one. Second, the federal government would take a minority, non-controlling equity stake in every recipient as a condition of the award.
The first feature answers a question the field has been asking since the CHIPS and Science Act passed in 2022: should the U.S. bet on a single integrated quantum manufacturer, or on an open foundry fabricating for a portfolio of qubit architectures? Commerce's LOIs propose funding both. IBM's standalone Anderon, which IBM frames as "America's first purpose-built quantum foundry" in its release, is proposed to receive $1 billion in CHIPS funds matched by $1 billion of IBM's own capital... a pure-play, single-modality model running superconducting qubits at scale on a 300mm wafer line at Albany NanoTech. GlobalFoundries' Quantum Technology Solutions, proposed to receive $375 million, is the open multi-modality model: a single foundry fabricating superconducting, trapped-ion, photonic, neutral-atom, silicon-spin and topological devices for an external customer list that includes PsiQuantum, Quantinuum, Diraq, Equal1, Quantum Motion, Microsoft and Google Quantum AI.
The second feature is a more visible and aggressive use of equity-based return mechanisms than in prior CHIPS-era deals. The CHIPS R&D Broad Agency Announcement already contemplated that awardees "may be required to issue equity, warrants, IP licenses, royalties, or revenue sharing," and the CHIPS Commercial Fabrication NOFO allowed loans and loan guarantees alongside grants. The quantum LOIs apply equity as a precondition across an entire $2 billion package. Per Commerce, the department "will receive a minority, non-controlling equity stake in each company as a condition for receiving the funds." Reuters and the WSJ both confirm the structure; D-Wave told the WSJ its entire proposed $100 million would be structured as equity, and Rigetti and Infleqtion are indicating comparable structures. GlobalFoundries disclosed that the government would hold "approximately one percent" of the company. IBM did not disclose its stake size, a gap the Standard noted in same-day coverage.
Commerce-led equity instruments at this scale are recent. The August 2025 Intel package converted $11.1 billion of pending and disbursed CHIPS funding into a 9.9 percent non-voting position. The November 2025 Vulcan Elements announcement proposed $50 million in CHIPS incentives to Vulcan Elements, alongside a separate $700 million conditional loan commitment from the Department of War's Office of Strategic Capital to ReElement Technologies. On a different track, the Pentagon's July 2025 MP Materials deal layered preferred stock and warrants on top of an offtake agreement, with a path to up to 15 percent. The CNBC equity-portfolio tracker catalogues the broader cross-agency pattern. The quantum LOIs extend Commerce's specific use of it from classical chip fabrication into frontier compute.
Anderon and GlobalFoundries Quantum Technology Solutions sit on opposite ends of the foundry-design axis. Read them together and the parallel funding decision is the structural move, not the headline dollar amount.
Anderon, per IBM's own framing, is "America's first purpose-built quantum foundry." The proposed $1 billion CHIPS award plus the $1 billion IBM-side investment would commercialize the 300mm wafer line IBM has been running at NY CREATES' Albany NanoTech Complex, where the Quantum Loon and Quantum Nighthawk processors are already fabbed. The LOI converts that internal capability into a foundry available to external quantum customers.
GlobalFoundries Quantum Technology Solutions answers the same problem from the opposite direction. Rather than a vertically integrated stack tied to one qubit architecture, GF is proposing to build a multi-modality production line at its Malta, New York fab, covering cryogenic CMOS, advanced packaging, and superconducting interconnects for the customer list above. The architecture is structurally close to the open multi-project wafer model the EU launched with its SPINS spin-qubit pilot line in April, with one important difference. SPINS is a €50 million imec-coordinated pilot line designed to move spin qubits from lab to fab. GF Quantum Technology Solutions would be a commercial 300mm production foundry funded at roughly seven times that scale and downstream of the lab-to-fab transition.
Anderon optimizes for vertical control over a single high-volume modality; GF Quantum Technology Solutions optimizes for portfolio coverage across a field in which the architectural winner is not yet identifiable. Commerce is proposing to fund both, mirroring what DARPA has been doing with its hybrid-qubits program: hedge across modalities because the path to a fault-tolerant quantum supercomputer is not yet settled.
The seven non-foundry recipients map cleanly onto the leading qubit modalities, with $100 million proposed for each except Diraq ($38 million). Atom Computing and Infleqtion cover neutral-atom systems, with Atom focused on tens-of-thousands-qubit integration and Infleqtion on optical systems and error correction. PsiQuantum covers photonic, with engineering challenges centered on high-temperature single-photon detectors and ultra-low-loss photonic packaging, the same modality that surfaced its capital-market story when Xanadu went public. Quantinuum covers trapped-ion. D-Wave and Rigetti cover superconducting from two angles: D-Wave on annealing-plus-gate-model architectures and advanced packaging, Rigetti on next-generation cryostat architecture and readout-electronics miniaturization. Diraq is the silicon-spin recipient, with a fab partnership at GlobalFoundries for the cryo-CMOS layer.
Topological qubits remain the least mature of the modalities GF would support, which is why Microsoft's endorsement of the GF foundry carries weight; the Majorana qubit readout result reported earlier this year is what gives topological a place on a 2026 manufacturing roadmap at all. The point of the modality spread is not that all will succeed at fault-tolerant scale. It is that Commerce appears unwilling to bet that any single one will. Gregg Bartlett, GlobalFoundries' chief technology officer, framed quantum in his company's release as "the next generation of supercomputing," which is also the right way to read the deal at the category level.
The instrument shift is the part of the package that matters beyond quantum. CHIPS Act 2022 awards were predominantly grants conditioned on construction milestones, workforce commitments and upside-sharing if returns exceeded expectations, with loans and loan guarantees available alongside. The August 2025 Intel deal restructured $11.1 billion of pending and disbursed CHIPS funding into a 9.9 percent non-voting common-stock position, per the company's Form 8-K. The quantum package extends that equity-based posture into R&D-stage frontier compute, with equity as a precondition rather than a back-end conversion. The pattern has not been litigated to settlement; Intel is currently facing a shareholder lawsuit over the structure of its deal, a precedent policy partners reading the quantum LOIs will be tracking closely.
For the supercomputing supply chain question, the more consequential read is federal balance-sheet exposure. The U.S. would become a minority shareholder in nine quantum companies before any of them has demonstrated a fault-tolerant device. The downside is direct, since equity in pre-revenue hardware companies can mark down; the upside is symmetrical, since the CHIPS R&D Office could ride a successful quantum supercomputer manufacturer in a way that grant-and-claw-back never permitted. Jacob Feldgoise at Georgetown CSET has flagged that trade through the Intel coverage. It is the structural change worth tracking regardless of which recipients reach scale.
The U.S. package lands roughly seven weeks after the EU's SPINS pilot line launch on April 3, and weeks after China formalized quantum as a "future industry" in its 15th Five-Year Plan. Read against those reference points, the May 21 LOIs document a third national-industrial-policy posture rather than the first one. The U.S. is following the EU's open-foundry posture by weeks, not years.
The EU's approach, under the Chips Joint Undertaking, is to fund six platform-specific quantum pilot lines coordinated by imec, each structured around open multi-project wafer access and standardized quantum process design kits to move research-grade fabrication into industry-grade processes. SPINS is funded at €50 million across 25 partners in nine countries. GF Quantum Technology Solutions, at $375 million, would be funded at roughly seven times SPINS, and unlike SPINS would be a commercial production foundry rather than a lab-to-fab consortium.
China's 15th Five-Year Plan, approved in March, identifies quantum among the country's "future industries." Secondary policy-tracker analysis places the financial commitment at roughly $17.5 billion across regional quantum venture funds, though figures of that specificity are not stated in the primary plan text. The policy mechanism is government procurement preference, manufacturing subsidies, and mandated application deployments - the same industrial-policy playbook Beijing has used on semiconductors, solar, EVs and 5G.
The May 21 package would commit the U.S. to the EU's open-foundry logic with an equity instrument neither the EU nor China has used. The U.S.-China Economic and Security Review Commission's November 2025 report recommended exactly this kind of consolidated industrial-policy move; the LOIs are arguably its implementation.
The package is a set of letters of intent. The Commerce announcement notes that the deals "still require final deal completion"; IBM's release says the Anderon launch "is subject to the negotiation and execution of definitive documents." None of these are awards yet, and none of the equity stakes are closed. Definitive documents, equity-issuance mechanics, and likely Treasury and SEC procedural sign-off for public-company recipients would determine how much of the package becomes binding and how much falls out in due diligence.
What the LOIs establish, even before final close, is a policy posture. The U.S. is proposing to fund both foundry models for the domestic quantum manufacturing base, hedged across several leading qubit modalities, and to structure the instrument so the federal balance sheet would ride the outcome. The dollar number is incidental to that. The shape of the deal is the move worth reading.
