Anderon, a company under IBM, said on September 16 that it has finalized a 1-billion-dollar award from the US Commerce Department under the CHIPS Act, money earmarked for a pure-play quantum foundry in the United States. The deal extends federal chip subsidies into a technology that has spent years promising more than it has delivered.
The award is notable less for its size, small by the standards of the CHIPS program, than for its target. The government has spent most of its chip money on advanced logic and memory fabs, the factories that make the processors inside phones, computers and data centers. Quantum computing has been a separate, experimental corner of the industry.
A pure-play quantum foundry would build the physical machines, the chips, the cooling systems and the control electronics, that quantum computers require, and sell them to the companies and labs writing quantum software. Anderon’s pitch is that the industry needs a manufacturer dedicated to quantum hardware, the way TSMC builds chips for others.
Quantum computing has followed a familiar arc: early excitement, then a long stretch of hard engineering, then skepticism about when the technology would produce anything useful. The largest technology companies, IBM among them, have kept funding the work on the theory that the eventual payoff will justify the years of investment.
IBM has been in quantum longer than most of its rivals. The company built early quantum processors, opened them to researchers through the cloud, and published a road map that stretches years into the future. Anderon appears to be the vehicle through which IBM is turning that research into a manufacturing business.
The CHIPS award, according to the announcement, will support research and development for the foundry rather than construction of a finished plant. The distinction matters: the government is paying for the hard early work of figuring out how to make quantum hardware repeatably, which is precisely the part the industry has struggled with.
Analysts said the award signals that Washington sees quantum as a national-security concern as well as a scientific one. Quantum computers, if they mature, could break the encryption that protects most digital communication, and governments have begun treating the technology as a strategic competition with China.
The funding also connects to the broader debate about where chip money should go. The CHIPS Act was sold as a way to rebuild American semiconductor manufacturing, and most of its funds have gone to established players making conventional chips. The Anderon award moves a slice of that money toward a technology that has no commercial market yet.
IBM has not said how much of its own money it will put beside the federal award, or when the foundry might produce its first machines. A person close to the company said the effort will proceed in stages, with the CHIPS money covering the research needed before manufacturing can begin.
The company’s quantum business has been a source of pride and a target of doubt. IBM has shown working processors and hosted them for researchers, but it has also watched rivals, Google among them, claim their own advances in the field. The foundry effort is a bet that the industry’s future will be built by specialists rather than by each lab making its own hardware.
For Washington, the award is an experiment in extending the industrial policy of the CHIPS Act to a field where the factories do not yet exist. The money is real; the machines are not. Whether the 1 billion dollars buys a working quantum foundry, or another chapter in a long research program, will not be clear for years.
The CHIPS Act itself is the backdrop. The program set aside tens of billions of dollars to rebuild American semiconductor manufacturing, and the government has spent it on fabs in Arizona, Ohio and Texas built by Intel, TSMC and Samsung. The Anderon award is among the first to reach a company whose product is quantum rather than classical computing.
Quantum’s commercial promise has been slow to arrive. The machines built so far are too small and too error-prone to run the calculations that would justify their cost, and the industry has survived on government and venture funding while researchers work on the error-correction problem. A foundry would not solve that problem, but it would standardize the hardware the research depends on.
IBM’s rivals have made their own bets. Google has claimed advances in quantum error correction, and a growing list of startups are trying to build the machines and the software. IBM’s decision to structure Anderon as a foundry suggests it sees the manufacturing layer, not the algorithms, as where the next decade’s value will be.
Analysts said the more immediate question is how the award changes the quantum industry’s structure. A dedicated foundry, if it works, would lower the cost for startups and universities that cannot afford to build their own hardware, the same role TSMC played for the chip industry a generation ago.
