Nobody Knows Who Wins Quantum. America Funded Everyone
Right now, somewhere, encrypted data is being copied and stored by people who can't read it yet. They're betting they will be able to later. Security people call this harvest now, decrypt later. It works because most of the encryption guarding your bank account, your medical records and government communication rests on a single assumption, that certain math problems take longer than a lifetime to solve.
A quantum computer would solve them in hours.
This isn't a hypothetical anymore. Federal agencies now have hard deadlines to move their most sensitive systems onto quantum-resistant encryption. The deadline is the end of 2030 for key exchange and the end of 2031 for digital signatures. Governments don't set dates like that for technology they think is science fiction.
So it's worth understanding what these machines actually are, because most explanations leave you more confused than when you started. An ordinary computer stores everything as bits, each one either a zero or a one, and it works through possibilities one at a time. A quantum computer uses qubits, which can be zero and one at once. That lets it hold a vast number of possibilities together and test them in a single sweep, arranged so the wrong answers cancel out and the right one survives.
The gap this opens is hard to overstate. Think of the number that guards your bank data, a few hundred digits long. To break it, a computer has to find the two prime numbers that were multiplied together to make it. The fastest ordinary computer would need longer than the age of the universe to work that out. A quantum computer of the right size would need hours.
This doesn't make it a faster version of your laptop. It makes it a different kind of machine, one that is extraordinary at a narrow set of problems and no better than an ordinary computer at everything else.
But that narrow set includes some of the most valuable problems there are, above all simulating molecules and materials. Ordinary computers are bad at it because nature itself is quantum, and cracking it is the path to better catalysts, batteries, fertilisers and drugs. Get that right and you can design medicines and materials that no existing machine can even model.
Here's the part everyone leaves out. A quantum computer is not a laptop or a server. Depending on the approach, it's a chandelier of gold-plated wiring hanging inside a refrigerator colder than deep space, or a room of lasers pinning individual atoms in place. A useful one needs somewhere between hundreds of thousands and millions of components, all behaving almost identically, all at the same time.
Which is why the first American machine of its kind is going up on the site of a steel mill.The South Works mill on Chicago's lakefront closed in 1992. The land sat empty for three decades. Last September, crews broke ground there on a quantum computing campus with PsiQuantum as the anchor tenant. I keep expecting quantum computing to look like a laboratory. It keeps looking like a construction site.
In May, the US government made that shift official, and almost nobody read the document properly. It put $2 billion into quantum computing. Two thirds of it went to companies that don't build quantum computers.That sounds like a mistake. It isn't. It's the clearest thing anyone has said about what's actually holding this technology back, and it came from the buyer with the best information in the market.
Here's what happened.
Since 2022, the CHIPS and Science Act has funded quantum the way governments usually fund science. Money to universities, to the National Science Foundation, to Department of Energy labs. Grants and papers.
The May money came from a different pocket of the same law. It came from the office set up to bring semiconductor manufacturing back to America. That office signed nine agreements worth $2.013 billion, wrote them as factory funding rather than research funding, and took a small ownership stake in every company as a condition of the cheque.
The switch matters more than the number. Research budgets fund a field. Factory budgets build an industry. Now look at who got paid.
IBM took $1 billion to set up a subsidiary that will manufacture quantum chips. GlobalFoundries took $375 million to build a plant that can make chips for five different kinds of quantum computer. GlobalFoundries is a contract manufacturer. It makes chips for other people and has no quantum computer of its own. Neither company is a quantum computing company, and together they walked away with 68 percent of the money.
The seven actual quantum computing companies split the rest. Atom Computing, D-Wave, Infleqtion, PsiQuantum and Quantinuum each got exactly $100 million. Rigetti got up to $100 million.
Read that again. Six identical cheques, the same number six times. And those six aren't building the same machine. Each has picked a different physical way to make a qubit. One traps individual atoms with lasers. One uses superconducting circuits chilled to near absolute zero. One uses single particles of light. One uses the spin of an electron in silicon. These are competing bets on the basic design, and at most one of them is right. Possibly none of them are.
So writing the same cheque six times isn't six separate bets on which approach is best. It's one bet that nobody can pick the winner yet. The government all but said so. Its official called the plan "a portfolio approach to strengthen and accelerate U.S. leadership across multiple quantum modalities at once."
Now read what these companies were actually paid to solve, because that's where the game gives itself away.
PsiQuantum's machine runs on single particles of light. Its money went to stopping that light from leaking away as it travels between components. Rigetti's chips sit inside a refrigerator, and its money went to shrinking the fridge and the tangle of wiring that feeds it. The rest were funded for the same kind of thing, packing parts more tightly, keeping materials consistent from batch to batch, controlling tens of thousands of components at once without any of them drifting.
Leaking light, tangled wiring, loosely packed parts, quality that drifts from one batch to the next. Every one of these is a problem you'd recognise from any factory floor. Not one of the seven agreements asks a company to figure out how quantum computing works. They ask it to figure out how to build the same thing twice.
The chip industry took fifty years and trillions of dollars to learn how to make ordinary processors reliably at volume. Quantum machines are harder, and you can't get there from a lab bench.Which is why the biggest cheques went to the factories. Every approach, whichever one wins, has to be built at scale. So factory money is the one bet in the whole programme that a breakthrough can't wipe out.
That changes what's worth paying for. Whether a company's physics will win is unknowable. Whether it can actually get its chips built is not. PsiQuantum is the one company where you can already see the answer. It makes its chips on the same GlobalFoundries line the programme just funded with $375 million. Its route to production isn't a hope. It's already paid for.
Two honest caveats. These are letters of intent, not contracts, and none of the money has actually moved. And no company anywhere has built a quantum computer that works reliably at commercial scale. Every approach on that list is still a research problem, now carrying a factory budget.
So here's where it lands.
Nobody knows which of these approaches will work. Not the companies, and not the government that just funded all of them. That's not a failure of analysis. It's the honest state of the technology, and the Commerce Department is the only player to have said so out loud, by writing the same cheque six times.
So it backed the category instead of a winner. Small money across every serious approach, and the real money into the one thing every approach needs no matter which turns out to be right, the ability to manufacture these machines at volume.That's a sensible way to fund something this early. It's also a signal about how seriously it's being taken. Governments don't build factories for technology they expect to stay in a lab.
The reason is in the stakes. The first working machine will break the encryption that banks, hospitals and governments rely on today. It will also design materials and medicines that no computer we currently have can model. Whoever gets there first can build what nobody else can model, and read what the world encrypted before it switched.
Nobody can say yet which company builds it. What's worth noticing is that America has stopped waiting to find out.
Bashar Aboudaoud
Managing Member, UpRound
Q3 2026 Live Webinar
Live Webinar: Thursday, September 17
You leave with three things. Where private valuations actually sit right now. Which sectors are repricing fastest. How I am reading prices into 2027.
20 minutes, then live Q&A. 20 slots only for UpRound readers.
