Spending ten billion dollars on something that most people still don’t fully understand gives rise to a certain kind of institutional confidence. That’s where IBM is at the moment; at the end of May, it filed an SEC disclosure announcing an investment in quantum computing that, over the next five years, would surpass any single commitment made by the private sector to the technology. The news wasn’t entirely unexpected. However, its magnitude, which coincided with the federal government’s decision to purchase stock in nine quantum companies, seemed to indicate a change. Washington has quietly and consciously begun to view quantum computing as an infrastructure priority rather than as a research curiosity.
The role of the government in this situation merits careful consideration. The Trump administration’s $2 billion investment in nine companies, of which IBM will receive about half through a new venture called Anderon, represents a fairly aggressive approach to federal tech policy. With its headquarters located in Albany, New York, Anderon would go on to become the nation’s first dedicated quantum chip manufacturing facility, producing about 300-millimeter wafers. IBM is matching the $1 billion in CHIPS Act incentives from the Department of Commerce with cash, intellectual property, and workforce support. Although it’s still unclear if a single foundry can significantly address the supply chain bottlenecks that have slowed the industry, the reasoning is sound: manufacturing has always been the place where aspirations and reality collide.
It’s worth taking a moment to examine IBM’s internal statistics. According to the company, it has already deployed over 90 quantum systems worldwide, which it claims surpasses the total amount disclosed by all other industry participants put together. More than 325 Fortune 500 companies, academic institutions, startups, and governmental organizations are now part of its quantum network. On IBM’s architecture, researchers from Oxford, ETH Zurich, Cleveland Clinic, and RIKEN have been conducting actual experiments, including modeling molecular systems, probing iron-sulfur clusters, and simulating protein structures. These accomplishments aren’t press releases. According to peer-reviewed findings published in journals such as Science and Nature Physics, quantum-classical hybrid computing is starting to provide value in specific but significant ways.
Fault tolerance is the technical challenge that everyone in this field is constantly discussing. Modern quantum systems are brittle; errors build up, qubits are disrupted by ambient noise, and computations fail before they are complete. IBM has set a lofty, perhaps even optimistic, goal for a large-scale fault-tolerant machine by 2029. Last year, Sundar Pichai of Google stated that it would still take five to ten years for quantum computers to be practically useful. Depending on what “useful” really means, both timelines might be accurate. In one version of this, fault tolerance is still just out of reach by 2029; in another, IBM’s investment compression causes the breakthrough to occur sooner than anticipated.
The reference architecture IBM unveiled in March is what distinguishes this moment from previous quantum announcements. Built on open frameworks like Qiskit, it is a technical blueprint for integrating quantum processors with GPUs and CPUs on-premises, in research facilities, and in the cloud. Developers now have a published, scalable model for how quantum computing fits into current infrastructure for the first time. Although it’s not as glamorous as revealing a new chip, it might have greater significance. Hardware milestones are rarely followed by practical adoption; instead, usable systems come first.
It’s easy to understand Washington’s underlying motivation. About $15 billion has been spent by China on quantum technology, of which $10 billion has gone through the Hefei National Laboratory. Policymakers have made it clear that quantum systems have the potential to undermine current encryption, alter drug development and logistics, and alter the distribution of computational power. The CHIPS Act funds, the government’s equity stakes, and the Albany foundry are all components of an industrial policy that, at least structurally, resembles Operation Warp Speed. Everyone in this field is silently wondering if it yields results that are comparable.
