Four experimental nuclear reactors have started self-sustaining chain reactions on U.S. soil in a single month, clearing a symbolic July 4 deadline set by the Trump administration and, backers say, pulling American nuclear power out of a decades-long slump. Whether the achievement leads to electricity on the grid is another question entirely.
The reactors — built by Antares Nuclear, Valar Atomics, Deployable Energy, and Aalo Atomics — each achieved what physicists call criticality, the point at which a fission chain reaction sustains itself. Antares crossed the line first at Idaho National Laboratory on June 4. Valar followed on June 18 at Utah’s San Rafael Energy Lab. Deployable’s Unity reactor went critical late on June 30, and Aalo’s Critical Test Reactor woke up in the early hours of Independence Day.
The count exceeded the target of three set by President Trump’s May 2025 executive order, which directed the Department of Energy to fast-track qualified test reactors at federal facilities. Energy Secretary Chris Wright has framed the sweep as evidence that the country can build new nuclear hardware in months rather than decades.
What actually happened inside the reactors
Criticality is a physics threshold, not a power plant switch. In a zero-power criticality test, engineers slowly load fuel and withdraw control rods until neutrons produced by fission trigger exactly enough new fissions to keep the reaction going. The reactor is alive but idling. It generates almost no heat and no electricity.
Kathryn Huff, former assistant secretary for nuclear energy, has noted that the milestone is easy to overread. A zero-power-criticality test can be achieved without making real engineering progress on fuel or design.
Still, the demonstrations validated four distinct approaches. Antares built a sodium heat-pipe-cooled microreactor fueled with high-assay low-enriched uranium in TRISO pellets. Valar’s Ward 250 uses TRISO fuel and helium coolant in a high-temperature gas configuration. Aalo’s design is sodium-cooled, graphite-moderated, and runs on standard low-enriched uranium dioxide. Deployable’s Unity, described in detail by POWER magazine, borrows from three different reactor lineages — water for moderation, helium for cooling, and off-the-shelf 4.95% enriched uranium dioxide fuel.
Startups measured in months, not decades
The companies themselves barely existed a few years ago. That any of them fueled and started a functioning fission reactor in that window is the part industry observers keep returning to.
Deployable’s timeline is the most extreme. The Houston-based company went from project kickoff to a delivered reactor, delivered fuel, and criticality readiness in roughly 150 days on what the company has described as a single-digit-million-dollar budget. Aalo’s Critical Test Reactor was built in 36 days after site work began at Idaho National Laboratory, then completed with modules trucked in from the company’s Austin factory.
The speed came partly from choosing not to reinvent supply chains. Deployable’s approach deliberately abandoned an earlier design that used specialized fuel and graphite moderation because it would be too expensive and too constrained by materials availability. The goal was to use what could be easily gotten.
Why the Pentagon and data center operators are watching
Microreactors are tens to hundreds of times smaller than the light-water reactors that supply about a fifth of U.S. electricity. Unity is designed as a 1-megawatt-electric unit that ships inside a standard 20-foot container. Aalo is aiming for 10 megawatts by 2027 to power an on-site data center.
The interest from hyperscalers is not hypothetical. Michael Goff, principal deputy assistant secretary in DOE’s Office of Nuclear Energy, told TechSpot that several pilot-program companies are already in discussions with data center developers, with commercial announcements expected within a few years. The administration has set a target of quadrupling U.S. nuclear capacity to 400 gigawatts by 2050.
Aalo counts Microsoft among its backers. Deployable is pitching Unity for national security bases, remote industrial sites, and maritime uses. Small reactors placed next to a data center could sidestep the transmission bottlenecks that currently make it hard to plug new AI workloads into the grid.
The gap between criticality and commercial power
Every serious voice in the sector has spent the past week pointing out how far these projects still are from selling electrons. Alison Hahn, former head of advanced reactors at DOE and now at the Nuclear Energy Institute, put it bluntly: the pilot program is for demonstration reactors, not commercial power to the grid.
Cooling systems still have to be added. Fuel supply is a real bottleneck — several designs depend on high-assay low-enriched uranium, or HALEU, which has no commercial U.S. producer today. And the Nuclear Regulatory Commission, not DOE, holds the authority to grant commercial licenses. A proposed Part 57 framework would create a faster path for microreactors with comparable risk profiles, but it has not been finalized.
Financing is another wall. DOE’s pilot program does not fund construction. Edwin Lyman of the Union of Concerned Scientists has called the program a performative exercise without federal financing attached, arguing that capital costs remain the biggest obstacle to actually deploying reactors.
The center-left think tank Third Way has been sharper. In a recent analysis of U.S. nuclear policy in 2026, the group argued that artificially accelerating project timelines is a short-term solution, not a long-term fix, and warned that federal focus on microreactors may be distracting from more than $17 billion in federal loans aimed at larger reactors that could deliver more gigawatts, faster.
The pilot program’s next moves
The Reactor Pilot Program, launched in June 2025, accepted 11 projects from 10 companies: Aalo Atomics, Antares Nuclear, Atomic Alchemy, Deep Fission, Last Energy, Natura Resources, Oklo (with two projects), Radiant Industries, Terrestrial Energy, and Valar Atomics. Four have now hit criticality. Oklo Isotopes is targeting first criticality for its Groves test reactor in Texas later this month.
Oklo’s larger Aurora-INL project — a 75-megawatt sodium-cooled fast reactor under construction at Idaho National Laboratory — is aimed at commercial operation in 2028. Deployable also plans to have commercial reactors operating by 2028, targeting speed-to-power under six months once designs are licensed.
DOE has also launched a follow-on effort, the Nuclear Energy Launch Pad, which opens roughly 2,000 acres at Idaho National Laboratory to private developers and offers a second track for projects on non-federal land. Deployable’s Unity was the first criticality achieved under Launch Pad.
What the milestone actually proves
The four criticalities settle a specific question: whether small teams with modest budgets can move a novel reactor design from concept to a sustained fission reaction inside a couple of years. The answer, for the first time in generations, is yes.
What the milestone does not settle is whether any of these designs will produce affordable, licensed, grid-connected power. That depends on fuel supply, financing, NRC review, and the boring engineering of turning a hot reactor core into a working power plant. Companies still need cooling loops, turbines, and safety cases that will hold up to commercial licensing rather than DOE authorization.
Related developments in nuclear safety and materials science are moving in parallel. Advances such as Tetris-inspired radiation detectors and 3D-printed tungsten reactor components are the sort of supporting technologies that a factory-built microreactor industry would depend on. Portable monitoring gear developed for nuclear security and arms control may also find a role as small reactors proliferate at commercial sites.
Goff, at DOE, has been careful to describe criticality as the first step on a longer path. If the industry stopped here, he said, the focus might be misplaced. The next 18 to 24 months — the fuel, the cooling systems, the NRC filings, the first customer contracts — will decide whether the July 4 deadline was a real turning point or an unusually well-timed press event.