Engineers in Nevada ran a microreactor core critical for four weeks at over 1,470 degrees, and the heat that made those temperatures came from electric elements rather than from the fission inside it

Inside a concrete bay in southern Nevada sits a steel vacuum chamber about the size of a kitchen appliance.

No pump runs anywhere in the building.

Inside the chamber, slim sealed tubes carry heat out of the assembly with nothing moving them but physics.

That means the chain reaction was self sustaining while the fission itself produced almost no measurable heat, and almost no fission products with it.

For four weeks the thing inside held a chain reaction.

It also produced almost no heat of its own.

That second sentence is the whole point, and it is the part the headlines dropped.

How a sealed tube replaces a coolant pump

Every conventional reactor built in the last eighty years has solved the same problem the same way. Push a fluid through the core fast enough to carry heat away, and keep pushing.

Lose the pump and you lose the flow, and losing the flow is how cores melt.

A heat pipe removes the pump from the question entirely.

Each pipe is a sealed, evacuated tube holding a small charge of sodium or potassium. At the hot end that metal boils, the vapor rushes to the cold end, gives up its heat and condenses, and a wick pulls the liquid back.

There is no valve to stick, no loop to rupture and nothing to switch on.

The physics is proven in spacecraft and electronics cooling. What has never been settled is whether that passive transport holds steadily enough inside a fueled core.

What was actually on the test stand

The reactor is called ZiaCore, and the stand it sat on is called Deimos.

That distinction matters, because plenty of coverage has mixed them up. Deimos is a reusable platform for criticality experiments at the National Criticality Experiments Research Center in Nevada.

ZiaCore is the thing being tested on it.

The core is moderated by zirconium hydride, a dense ceramic that slows neutrons efficiently at high temperature and has been used in research reactors since the 1950s.

Its fuel is low enriched uranium dioxide, with the design written to accept high assay fuel later if domestic supply arrives.

Most of the hardware, including the heat pipes and the moderator, was built in house on a budget of 5.5 million dollars over three years.

What four weeks in Nevada actually measured

The campaign ran across April and May of 2026 and was announced at the end of July.

The core was brought above 1,470 degrees Fahrenheit, and it was brought there by custom electric heaters wrapped around the assembly.

Then it was taken critical at essentially zero power.

That means the chain reaction was self sustaining while the fission itself produced almost no measurable heat, and almost no fission products with it.

What the team collected was core physics at operating temperature, above all the temperature coefficient of reactivity, which is the number that says whether a core damps itself as it heats.

The design target is roughly the output of a large diesel generator, running eight years without refueling. Neither of those numbers was demonstrated here.

What a zero power test does not tell you

It does not tell you the heat pipes move fission heat, because there was none to move.

It does not tell you how the assembly behaves through thermal cycling across years, or what the hydride moderator does as it ages.

And it says nothing about power density, which remains the real gap. A microreactor core at this scale can run a remote drilling site or a forward base. It falls well short of a utility.

None of that makes the result small. Confirming that a core’s physics match prediction at temperature is the step that has to come first, and the measured coefficients are what a licensing case is built from.

The claim worth being careful with is priority. The narrow version holds, a heat pipe core with this moderator and this fuel reaching criticality for the first time.

The broad version does not, because the design joins several concepts arriving at the same milestone.

What else went critical in the same twelve months

The field is crowded now, and the order is worth having straight.

A graphite moderated core with coated particle fuel reached zero power criticality at the same Nevada facility in November of 2025.

A privately developed sodium heat pipe reactor went critical at a national laboratory in Idaho in June of 2026, billed as the first privately built non light water reactor to do so in over forty years.

A commercial heat pipe design followed in Nevada in August.

All of them are physics experiments rather than power plants, and all of them are waiting on the same bottleneck, which is domestic enrichment and coated particle fuel at volume. That constraint shapes every advanced build, from molten salt work in Tennessee to borehole pilots in Kansas.

The chain reaction has been confirmed. What has not been confirmed is everything downstream of it.