The Story
On May 21, the U.S. Commerce Department put out a list of nine companies it plans to hand a combined $2.013 billion to. The headline number is big, but the structure is the part worth slowing down for. This isn’t a research grant in the old sense. In exchange for the money, the federal government gets a minority, non-controlling equity stake in every single company on the list. Washington is buying shares.
The money comes from the CHIPS and Science Act — the same 2022 law everyone associates with semiconductor fabs — routed through the CHIPS Research and Development Office. So far that law has mostly funded chip plants in Arizona and Ohio. This is the moment it pivots toward quantum computing, and it does so with a twist on how the government usually writes checks.
Here’s the breakdown. The biggest slice, $1 billion, goes to IBM. Not to IBM directly, exactly — IBM is spinning up a new subsidiary called “Anderon,” headquartered in Albany, New York, and pitched as America’s first purpose-built quantum foundry. A foundry, in the chip world, is the factory that physically makes the wafers. Anderon’s job is to manufacture quantum-grade superconducting wafers at scale — the raw silicon-and-metal substrate that superconducting qubits are etched onto. GlobalFoundries gets $375 million to build a second foundry, this one deliberately architecture-agnostic: superconducting, trapped-ion, photonic, topological, silicon spin. The government clearly doesn’t want to bet the supply chain on one qubit design.
The other $625 million or so is spread across seven quantum computing companies, most getting roughly $100 million each: Atom Computing, D-Wave, Infleqtion, PsiQuantum, Quantinuum, and Rigetti. The Australian startup Diraq, which works on silicon spin qubits, gets up to $38 million. It’s a deliberately diversified portfolio. A Commerce official put it plainly — this technology “could take years,” so they’re spreading the bets rather than crowning a winner.
The market noticed immediately. Shares of the publicly traded names — D-Wave, Rigetti, GlobalFoundries — jumped, with some quantum-linked stocks up between 7% and 21% in premarket trading. GlobalFoundries closed up around 9%. But it wasn’t a clean rally for everyone. Infleqtion reportedly fell almost 16%. When the government picks winners, the act of picking also reprices everyone — and a $100 million check that comes attached to a dilutive equity stake isn’t unambiguously good news for existing shareholders.
That equity condition is the real story. Commerce framed it as a way to “enhance the return for the U.S. taxpayer” — if quantum pays off, the public gets upside, not just a thank-you note. And this isn’t a one-off. The administration has already taken sizable positions in Intel and in the rare-earth miner MP Materials. Quantum is now the third sector where the U.S. government has moved from subsidizing companies to owning a piece of them. Call it the “equity-for-incentives” model. It’s becoming a pattern, not an experiment.
One quick reality check on the technology itself, because the word “quantum” gets thrown around loosely. None of these companies has a fault-tolerant, utility-scale quantum computer today. What exists right now are noisy machines with somewhere between dozens and a few hundred usable qubits — impressive lab instruments, not yet tools that reliably beat a classical supercomputer on a problem anyone cares about. The whole point of the funding language — “utility-scale, fault-tolerant” — is that it’s describing a destination, not the current address. This money is meant to shorten the road.
The Takeaway
The instinct is to read this as a story about quantum computing. I think it’s more useful to read it as a story about how the U.S. government now does industrial policy — and quantum just happens to be the latest test case.
Look at the pattern across what this blog has been tracking. When we covered OpenAI’s large-scale infrastructure expansion and the Samsung–NVIDIA “AI megafactory,” the through-line was that frontier computing has gotten too capital-heavy for normal market mechanics to handle alone. Quantum is the same problem, earlier on the curve. A quantum foundry is brutally expensive to stand up, the payoff is years out, and the science might still take an unexpected turn. That’s exactly the kind of bet private capital underprices — too long, too uncertain. So the government steps in. The novelty isn’t the government spending money on hard science. It’s that this time it wants equity. The 2022 CHIPS Act handed out grants; the 2026 version of CHIPS spending wants to be a shareholder.
The smartest thing in the package, honestly, is the refusal to pick a qubit. GlobalFoundries’ foundry is explicitly built for five different modalities. The seven computing companies span superconducting, trapped-ion, neutral-atom, photonic, and silicon-spin approaches. Nobody actually knows which of those wins — it’s a genuinely open scientific question — and the structure of this deal admits that out loud. Compare it to the AI hardware race, where we wrote about Google’s “Ironwood” TPU and Qualcomm muscling into the data center: there, the architecture question is mostly settled and the fight is execution. Quantum isn’t there yet. Funding the whole spread is the correct move when you don’t know the answer. It’s also the expensive move, and some of these $100 million bets will simply not work out. That’s not a flaw in the plan. That’s the plan.
There’s a thread back to something we covered in February too — the demonstration of quantum-safe data transmission. That story was about defense: building encryption that survives the day a quantum computer can break today’s codes. This story is the offense side of the same coin — actually trying to build that computer, on home soil, with a domestic supply chain so the wafers aren’t coming from somewhere else. A government that’s simultaneously hardening its cryptography and financing the machine that threatens it isn’t being contradictory. It’s hedging both ends of a transition it considers inevitable.
So what should you actually watch? Two things. First, whether the letters of intent become signed, funded deals — these are intentions, not wires transferred, and CHIPS money has a history of moving slowly. Second, and more interesting: whether “equity-for-incentives” keeps spreading. Intel, MP Materials, now nine quantum companies. If the next strategic sector — biotech, advanced batteries, whatever it is — gets the same treatment, then this announcement was never really about quantum. It was about a government quietly redefining itself as a venture investor, with public money and a public balance sheet. The qubits are the easy part to report. The shareholding state is the part that’ll still matter in ten years.
Photo: Ilya Pavlov / Unsplash
댓글 남기기