by Barry Manz, Editor
The U.S. government has decided that quantum computing is no longer a research line item. The Commerce Department said it would take equity stakes in nine quantum-computing companies tied to roughly $2 billion in funding, a move it framed as building a domestic quantum supply chain and countering China in advanced computing. The money is tied to CHIPS Act grants, and Commerce Secretary Howard Lutnick pitched it as spurring “a new era of American innovation.”
The dollar figures are the headline, but for the RF and microwave industry, the destination of the money is the story, because the machines this funding is meant to manufacture are, beneath the quantum mystique, some of the most demanding microwave systems ever built.
Two Fabs, Not Nine Grants
The largest share goes to IBM, and it is not a simple transfer. IBM is launching a company called Anderon in New Albany, NY, billed as America’s first dedicated quantum chip manufacturing facility, contributing $1 billion in capital, intellectual property, assets, and workforce, and attracting additional investors as the company scales. The government’s stake in Anderon was not disclosed.
GlobalFoundries takes the next-largest piece, $375 million to build a domestic factory capable of producing components for several different types of quantum machines, through a new business it has launched called Quantum Technology Solutions, with the government taking an equity stake of about 1%.
Strip away the politics, and what remains are two new manufacturing operations aimed specifically at quantum hardware, one from a systems company that has run its own quantum program for years, the other from a contract chipmaker that knows how to scale a process. The smaller awards round out the field across competing approaches. D-Wave, Rigetti, and Infleqtion each receive about $100 million, and Diraq receives up to $38 million to address key technical hurdles. Those five, plus IBM and GlobalFoundries, account for seven of the nine; the remaining names were not all disclosed, though Quantinuum was reported as a likely recipient in the $100 million tier.
Quantum As Microwave Instrument
Building a factory for quantum hardware is not the same problem as building one for logic or memory, and the reason will be familiar to anyone who designs microwave subsystems for a living. A superconducting quantum processor is only the cold endpoint of a large and intricate microwave instrument. The qubits themselves resonate in the neighborhood of 4 to 8 GHz, and every operation performed on them is a precisely shaped microwave pulse. Single-qubit gates are vector-modulated bursts a few tens of nanoseconds long, with amplitude, phase, and frequency controlled tightly enough to hold gate fidelities above 99.9%.
Generating them means a room-temperature rack of arbitrary waveform generators, IQ mixers, local oscillators, and vector signal sources, one chain per qubit, with the phase noise and spurious performance of every component folding directly into error rates.
The signal path from that rack to the chip is where engineering turns punishing. Each control line runs down through a dilution refrigerator across several temperature stages, and at each stage it must be attenuated, typically 20 dB at the 4 K plate and more below, to strip the room-temperature thermal photons that would otherwise wash out a qubit sitting near 10 millikelvin.
That demands cryogenic attenuators with stable, well-characterized performance at temperatures for which their datasheets were never written. Readout runs the other direction and is no easier. The dispersive signal coming off a resonator is a handful of microwave photons, so it has to pass through ferrite isolators and circulators that protect the qubit from amplifier back-action, then into a near-quantum-limited traveling-wave parametric amplifier, then a HEMT low-noise amplifier at 4 K, before it is warm and strong enough to digitize. Holding the whole chain phase-coherent across that thermal gradient is the actual problem.
Then there is the wiring. Every qubit needs control and readout lines, and the count climbs with the machine. Today’s processors already thread hundreds of semi-rigid and superconducting NbTi coax lines through the cryostat, and the path to thousands or millions of qubits does not survive a linear scaling of that approach. The field is pushing toward frequency-division-multiplexed readout, cryogenic microwave switching, and integrated control electronics, which have moved closer to the chip precisely because the interconnect, not the qubit, is becoming the bottleneck.
A fab dedicated to quantum chips solves one layer of the problem, the device itself, while implicitly committing the country to scaling everything wrapped around it: the cryogenic attenuators and amplifiers, the ferrite components, the high-density microwave interconnect, the filters, and the room-temperature signal generation. That supporting hardware is specialized, low-volume, and today sourced from a thin bench of suppliers. For the RF and microwave industry, that is the more durable signal in Thursday’s announcement than any single grant.
The Model and the Argument
The financing approach is itself a departure. The deals convert traditional grants into ownership positions, mirroring the Intel stake the administration took last year. The reaction split along familiar lines. Supporters call it strategic capitalism, arguing that taxpayer-backed high-risk technology should capture the upside if the bets pay off. Critics see government equity in private firms as a distortion that blurs the line between regulator and shareholder. The market voted with enthusiasm: IBM rose more than 7%, and GlobalFoundries, Rigetti, D-Wave, and Infleqtion climbed between 10% and 20% on the news.
The strategic rationale is straightforward. Quantum computing sits on the short list of technologies where a clear lead carries national-security weight, from cryptography to materials and sensing, and the administration is treating domestic fabrication capacity as too important to leave to private timelines. The CHIPS Act lineage makes the intent plain: this is semiconductor industrial policy extended to the next contested process technology.
The Open Question
What the government has bought is a diversified position across the field rather than a wager on a winner. Spreading capital over superconducting, trapped-ion, annealing, and neutral-atom approaches hedges against backing the wrong architecture, which is prudent given that no one yet knows which will scale to fault tolerance first. It is worth noting that the microwave dependency does not disappear if superconducting qubits lose the race.
Trapped-ion and neutral-atom machines trade some of the microwave control burden for laser systems, but they still ride on RF trap drives and low-noise microwave chains, so a domestic quantum buildout is a microwave-component buildout under any of the architectures on the table. Whether the policy reads as sensible diversification or as the government running ahead of the science will not be clear for years. What is clear now is that quantum has crossed from the lab into industrial policy, and the microwave hardware that powers these machines has just secured a federal sponsor.