DOE’s $215 Million Q Competition Puts a Price Tag on Fault-Tolerant Quantum Computing – and a Deadline

September 17, 2026 – The U.S. Department of Energy announced the Quantum Genesis Q Competition, a milestone-based initiative with up to $215 million in planned funding to demonstrate the first fault-tolerant, scientifically relevant quantum computers.

The competition, sponsored by the Office of Advanced Scientific Computing Research within DOE’s Office of Science, invites for-profit domestic entities to submit proposals for quantum computers with at least 100 logical qubits capable of performing hundreds of millions of fault-tolerant operations alongside scientific demonstration programs.

The initiative is structured in two phases. Phase I provides fixed awards of up to $1.5 million per awardee for early milestones – $250,000 on approval of a technical and validation plan, and the remaining $1.25 million on delivery of a validated prototype. Phase II pools $100 million for awardees that demonstrate a first-generation scientifically relevant quantum computer with at least 100 logical qubits, with two additional bonus pools of $50 million each for demonstrations reaching 150 and 200 logical qubits. The DOE’s solicitation provides for each incentive pool to be divided equally among qualifying awardees following a planned September 2028 evaluation.

In a parallel effort, the DOE issued a $45 million Quantum High-Performance Computing Validation and Verification Testbed Lab Call. This separate program draws on DOE’s National Laboratories to independently characterize and validate competing systems across the full computing stack – from physical hardware and quantum gates to logical architectures, algorithms, and classical control systems.

Of the up to $215 million in total planned funding, $2.5 million is in fiscal year 2026 dollars. All outyear funding is contingent on congressional appropriations.

DOE anticipates selecting three to ten awardees and may select all, some, or none. The RFA leaves the qubit modality to applicants. Applications are due October 19, 2026, with a virtual informational webinar scheduled for September 25 at 1:00 p.m. Eastern.

The Q Competition implements the Quantum Genesis initiative that DOE announced on June 23, 2026, one day after President Trump signed Executive Order 14413, Ushering in the Next Frontier of Quantum Innovation. That order established the Quantum Computer for Application Development and Discovery Science (QC-ADDS) Effort. It also directed a national effort to deliver at least one fault-tolerant quantum computer to a DOE facility.

The competition also builds on the Office of Science Advisory Committee’s report on a Path to an Integrated Quantum Future, released the same week, and the earlier Blueprint for DOE Quantum Supercomputing.

DOE Under Secretary for Science Darío Gil, who oversees the initiative, described the competition as building “a new era of computational power for the nation through innovative public-private partnerships.”

My Analysis

Milestone Payments, Equal-Split Pools, and a $45 Million Testbed

The headline number is $215 million, but the competition’s design tells me more than the dollar figure.

Start with Phase I. Each awardee receives $250,000 when DOE approves its technical and validation plan. The remaining $1.25 million arrives only on delivery of a validated prototype – roughly a year into the contest. DOE structured these as procurement milestone payments instead of a research grant, and it kept the amounts intentionally modest. By paying only $250,000 up front, DOE selects for companies that already have substantial capital and a viable path to fault tolerance, not those who need federal money to start.

Phase II uses the same approach. The $100 million pool is divided equally among everyone who demonstrates 100 logical qubits at the planned September 2028 evaluation. If five companies clear the bar, each gets $20 million. If one does, it takes the full pool. DOE adds two $50 million bonus pools to reward further scaling, one at 150 logical qubits and one at 200. A company that reaches 200 logical qubits collects shares of all three pools. Because the split is equal and the evaluation date is fixed, a company that reaches 100 logical qubits early gets the same share as one that qualifies at the September 2028 evaluation.

The $45 million V&V testbed is a separate program. DOE’s National Laboratories will develop a rigorous framework and software tools for assessing quantum-computing performance across application, algorithmic, logical, and physical-qubit levels. In a field where vendor benchmarks are often internal, unaudited, and conveniently favorable, building independent verification infrastructure through the national lab system is a significant structural decision. If successfully deployed, the testbed gives DOE a common evidentiary basis for comparing claims that are otherwise difficult to evaluate across hardware modalities.

What a 100-Logical-Qubit Machine Can and Cannot Do

The Q Competition targets what DOE calls a “scientifically relevant quantum computer” – a system that can tackle currently intractable problems in chemistry, materials science, and plasma and high-energy physics. DOE will specify the scientific workflow at the start of the program; the RFA notes that the first-generation demonstration need not operate at classically intractable scales.

The press release describes “hundreds of millions of fault-tolerant operations,” but the RFA itself counts hard operations separately from the total. These are typically non-Clifford operations, which are the expensive ones in any error-corrected architecture. The first-generation target in the RFA is on the order of 10⁵ hard fault-tolerant operations. That is a more precise and more demanding metric than a raw operation count.

A machine with 100 to 200 logical qubits executing scientific workloads at that level of hard-operation throughput would be a milestone in quantum error correction. It would demonstrate that the engineering problems of syndrome extraction, below-threshold scaling, and real-time decoding have been solved at a meaningful scale.

A system at this scale is not a cryptographically relevant quantum computer. The best current resource estimates for factoring RSA-2048 – the benchmark that defines when current public-key cryptography breaks – have come down sharply. Gidney’s May 2025 paper reduced the requirement to fewer than one million noisy physical qubits under a surface-code architecture, assuming 0.1% gate error rates and specific timing parameters. Related resource-estimation work puts the abstract logical-qubit requirement at approximately 1,400 under those assumptions. Earlier estimates ran into the tens of millions of physical qubits; the compression reflects advances in magic state cultivation and algorithmic fault tolerance, with qLDPC codes promising further reductions. But 100 to 200 logical qubits is still several-fold to more than an order of magnitude below the cryptanalytic threshold.

So the Q Competition affects the path to a CRQC only indirectly, through the credibility of the 2028 fault-tolerance timeline. If multiple companies demonstrate 100 logical qubits with verified fault-tolerant operations by September 2028, closing the engineering gap to a cryptographically relevant machine means scaling an error-correction architecture already shown to work – a large but now better-defined engineering problem, with open questions in logical error rates over longer computations, magic-state production at cryptanalytic throughput, decoder speed, and manufacturing yield. That is a qualitative shift in risk assessment for anyone responsible for cryptographic infrastructure.

From Executive Order to RFA in 87 Days

On June 22, 2026, President Trump signed two executive orders. Executive Order 14413 directed a national effort to build a quantum computer capable of enabling scientific discovery, with delivery to a DOE facility. The other, Executive Order 14412, set binding PQC migration deadlines for federal high-value assets and high-impact systems – key establishment by December 31, 2030, digital signatures by December 31, 2031.

On June 23, DOE announced the Quantum Genesis initiative. On September 17, DOE published the Q Competition RFA. Eighty-seven days from executive order to a competition with dollar amounts, milestone definitions, and application deadlines.

That is fast for federal procurement, and I suspect Darío Gil’s background is part of the reason. Gil spent more than two decades at IBM, the last several as Senior Vice President and Director of IBM Research, where he oversaw one of the largest quantum computing programs in industry. He was confirmed as Under Secretary for Science in September 2025. He brings both the technical credibility and the operational tempo to run a competition like this at speed.

Two Federal Deadlines Three Days Apart

Two federal quantum deadlines fall three days apart. Applications for the Q Competition are due October 19, 2026. Federal agencies must submit their PQC migration plans to OMB and the Office of the National Cyber Director by October 22, 2026 – three days later – under OMB Memorandum M-26-15.

The two instruments address different problems: one accelerates quantum computing capability, the other accelerates defense against it. The administration is treating quantum computing as both an opportunity to seize and a threat to manage, in the same budget cycle and with parallel urgency. I have argued before that debating when exactly a CRQC will arrive is close to irrelevant, because regulators, insurers, and procurement authorities have already set their own dates. The Q Competition adds another dated obligation to that growing roster.

What This Means for the CRQC Timeline

The Q Competition does not change my assessment of when a CRQC becomes feasible. Resource estimates move with algorithmic research and error-correction engineering.

What the competition does is stress-test the 2028 fault-tolerance claim. If the V&V testbed verifies that multiple vendors can sustain 100 logical qubits through hundreds of millions of fault-tolerant operations by September 2028, then the nine capabilities in my CRQC Quantum Capability Framework shift from research demonstrations to engineering-at-scale. Syndrome extraction, below-threshold operation, and real-time decoding all have to work together in one system that runs real scientific workloads continuously.

The distance from 100 logical qubits to the approximately 1,400 needed for RSA-2048 factoring is large. But if the error-correction architecture is proven at 100, the path to 1,400 becomes more credible – not easy, but defined. The resource estimates show the trend: Gidney and Ekerå estimated roughly 20 million physical qubits in 2019 under surface-code assumptions. Gidney’s May 2025 paper, incorporating magic state cultivation, reduced that to fewer than one million. Preliminary analyses applying qLDPC and other high-rate codes suggest further reductions, though those estimates remain architecture- and assumption-dependent and should not yet be treated as a settled floor. Each step narrows the gap between what the industry can build and what a cryptanalytically relevant computation would require.

For organizations planning their PQC migration, the Q Competition sets a September 2028 checkpoint for fault-tolerant quantum computing, and their migration has to be finished before a CRQC arrives. A credibly funded, independently verified, milestone-gated government program to build a 100-logical-qubit machine by 2028 makes the “quantum is decades away” objection harder to sustain by the month.

What the Q Competition Changes

DOE has turned a government aspiration into a procurement instrument with dated milestones, independent verification, and enough money to attract serious entrants. The machines it targets, at 100 to 200 logical qubits, are too small to break RSA-2048, but one of them may prove the engineering architecture for a machine that can. Boards, regulators, and procurement officers act on a running, independently verified fault-tolerant machine more readily than on a theoretical case for fault tolerance.