The U.S. Government Is Buying Quantum Stock Before the Deal Is Done

The U.S. government is writing checks it cannot yet cash — and taking stock in return.

The Department of Commerce announced this week nine non-binding letters of intent totaling $2.013 billion in CHIPS Act funding across quantum hardware and manufacturing companies. IBM leads the package with up to $1 billion in CHIPS dollars alongside $1 billion in company money to build a standalone 300-millimeter quantum wafer foundry in Albany, New York. GlobalFoundries receives $375 million for a multi-modality cryogenic CMOS foundry in Malta, New York. The remaining seven awards range from $38 million (Diraq) to $100 million each for Atom Computing, D-Wave, Infleqtion, PsiQuantum, Quantinuum, and Rigetti.

Here is the part most coverage is skipping: for every recipient, the DoC takes a minority, non-controlling equity stake as a condition of funding. And the awards are LOIs — non-binding at this stage. Washington is buying shares in unproven quantum companies before final terms are locked, before due diligence is complete, before a single qubit is manufactured at scale. The full terms of each LOI are available at NIST.gov.

This is not a grant program with strings attached. It is a shareholder arrangement dressed as an industrial policy announcement.

The equity mechanism is not incidental. It means taxpayers own a piece of nine companies whose core technology remains largely unproven beyond controlled laboratory conditions. Fault-tolerant quantum computing — the kind that would justify an $850 billion economic value estimate by 2040, as IBM cites from Boston Consulting Group — does not yet exist at commercial scale. The companies receiving money here are not manufacturing partners for classical chips. They are ventures betting on a timeline that has slipped repeatedly across two decades of optimistic projections.

IBM's $1 billion award is the largest single bet. The Albany foundry, if built, would be America's first purpose-built 300mm quantum wafer facility — a genuine manufacturing capability gap the country currently lacks. That part of the story is real and worth watching. But IBM itself acknowledges fault-tolerant quantum computing as a 2029 target. The LOI structure means the government is taking equity in that timeline before the foundry exists, before the process is qualified, before anyone can verify whether 300mm quantum wafers behave anything like their 200mm predecessors.

GlobalFoundries' $375 million is more grounded in existing manufacturing infrastructure. GF already operates at scale in Malta and has a working relationship with Quantinuum, which is also on the receiving end of this package. Quantinuum separately announced partnerships with GlobalFoundries and Monarch Quantum. The circularity is notable: the DoC is funding two companies that are also each other's manufacturing partners, and taking equity in both.

Five qubit modalities appear across the nine recipients — neutral-atom, silicon-spin, superconducting, photonic, and trapped-ion. That breadth suggests the government is hedging rather than picking winners, which is reasonable industrial strategy. It is also an admission that no single modality has demonstrated decisive advantage at scale. If one approach were clearly winning, a focused bet would make more sense than $2 billion spread across the field.

The equity stake structure is the revealing detail. A non-controlling government shareholding in early-stage quantum ventures is not standard procurement. It is a signal that Washington believes in these timelines badly enough to get in early — and to share in upside if they prove right. That is either prudent futurism or a $2 billion bet on a future that keeps not arriving.

The LOI stage means all of this is conditional. Final terms, detailed milestones, and enforceable commitments come later. What the announcement establishes right now is the government's intent to be a minority owner in nine quantum companies before anyone has proven these businesses can deliver.

That is either a bold industrial bet or an expensive way to buy optionality on technology that remains perpetually five years away. The equity paperwork will tell us which.