.png?w=1024)
The U.S. Department of Energy wants quantum computers judged by the science they deliver, not the qubits they count. A federal advisory committee has now laid out a three-phase plan to get there, and DOE paired it with a competition worth up to $215 million for private companies.
The plan, titled "Path to an Integrated Quantum Future", came from a Scientific Computing Advisory Committee subcommittee chaired by Anna Grassellino, Fermilab's chief technology officer. The quantum computing roadmap targets a scientifically relevant, error-corrected quantum computer by 2028. DOE unveiled it on Sept. 17, and Fermilab highlighted it on Sept. 25.
It signals where federal money will likely flow. The catch is that the plan is less specific than the competition built around it.
What Is in DOE's Three-Phase Plan?
Phase | Timeline | What happens | Aim |
| I: Quantum Grand Challenges | 2026 to 2028 | Competitions pair national labs, universities, and industry. | Scientifically relevant, error-corrected capability by 2028. |
| II: DOE Quantum Computing User Facility | After Phase I | An open, collaborative scientific instrument. | Timing, scale, and scope subject to "careful evaluation". |
| III: Integrated Quantum Future | 2030 and beyond | Quantum co-processors join DOE's AI and supercomputing systems. | Quantum computing increasingly integrated across DOE science. |
Sources: SCAC report, Fermilab.
Supratik Guha, a University of Chicago professor of molecular engineering, served as vice chair. Its science targets include catalyst chemistry, superconducting materials, fusion-relevant materials, and the quantum field theories behind particle physics.
The plan is also technology neutral. The report cites broad consensus that it is premature to converge on a single technological pathway. DARPA is running its own comparison, and evaluators can now test Microsoft's Majorana 2 chip hands-on at a University of Maryland facility.
How the Quantum Genesis Q Competition Fits
Darío Gil, DOE's under secretary for science, wrote, "Our goal is not simply to build the largest quantum computer; it is to solve problems that are otherwise completely intractable."
The Quantum Genesis Q Competition is open to private companies, with final applications due Oct. 19. Its structure:
- Phase I: Up to $1.5 million per awardee for early milestones
- Phase II: A $100 million pool for awardees that demonstrate a system with at least 100 logical qubits
- Bonus pools: Two $50 million pools, at 150 and 200 logical qubits
- Testbed call: A separate $45 million verification and validation call for national labs
The money follows $300 million in CHIPS Act awards that Commerce finalized in September for PsiQuantum, Quantinuum, and Rigetti. Those funds support manufacturing and scaling. DOE pays for demonstrated logical qubits.
The structure makes the distinction important: the roadmap defines the scientific destination, while the competition uses logical-qubit milestones to decide where near-term funding goes.
What eWeek Found: The Roadmap Preaches Utility, but the Payout Counts Qubits
Gil wrote that "success must be measured by scientific utility." The competition's largest rewards, however, turn on logical qubit counts. That is a hardware metric. The testbed call may be where utility gets checked, but it is not yet clear how DOE will score a result as scientifically relevant.
The numbers also sit far apart. The competition's minimum is 100 logical qubits. On page 23, the report says "approximately 1,000 logical qubits and sustained fault-tolerant operation" would enable a user facility after 2028.
On page 32, it says the facility's timing, scale, and scope "require careful evaluation as quantum technologies and scientific applications mature." Phase II therefore looks like a judgment call.
Three things will show whether DOE’s roadmap is working: which companies enter the competition, how national-lab testbeds define and verify “scientifically relevant” results, and whether any team can demonstrate useful fault-tolerant performance by 2028 rather than simply hitting a qubit threshold.
The roadmap’s central test is therefore not whether quantum computers get larger. It is whether DOE can turn hardware progress into scientific capability that justifies moving from experiments to a shared user facility.
Other news: French supercomputer maker Bull has doubled production at its Angers plant as Europe expands regional AI and high-performance computing capacity, although memory remains a major supply-chain dependency.