Xcimer Banks Another $30.5 Million

Two beamlines against one hundred and ninety-two. That is the comparison Xcimer Energy keeps making between its own architecture and the National Ignition Facility, and it is an optics argument before it is a fusion argument. On 6 October the Denver company said the US Department of Energy had finalised an agreement for up to $30.5 million in additional funding through its Milestone-Based Fusion Development Program, money paid out only as independently verified milestones are completed. It brings Xcimer's total potential milestone funding to $39.5 million and its total non-dilutive federal funding, including a $4 million ARPA-E award in 2024, to $43.5 million.

Set against the roughly $200 million Xcimer has raised privately — from Breakthrough Energy Ventures, Lowercarbon Capital, Emerson Collective and Prelude Ventures — the new award is not what is keeping the lights on. Set against a programme whose entire congressional authorisation under the CHIPS and Science Act runs to $415 million through fiscal 2027, shared across eight companies, it is a meaningful share of a small pot. What it buys, mostly, is validation that someone outside the company has checked the work.

The Milestone-Based Fusion Development Program was authorised in the Energy Act of 2020, announced by DOE in September 2022 and awarded to eight companies in May 2023. Its design is borrowed from NASA's commercial crew contracts and it inverts the usual research-grant arrangement: companies negotiate both technical and commercialisation milestones with the department, cover more than half the cost themselves, and are paid federal money only after an independent expert review confirms a milestone is done. Xcimer's new award covers the programme's second budget period.

Darío Gil, the department's Under Secretary for Science, framed the award in industrial-policy terms, saying that "America invented laser fusion, and America should lead its commercialization". Earlier in 2026 DOE signed off on Xcimer's preconceptual design and technology roadmap for Athena, the company's reference architecture for a commercial plant. The stated targets are electricity on the grid in the mid-2030s and a first commercial plant in 2035.

Inertial fusion is, from a procurement point of view, an optics problem with a plasma attached. NIF's 192 beams each terminate in large-aperture fused silica and KDP components that have to survive nanosecond pulses at ultraviolet wavelengths, and the laser-induced damage threshold of those final optics is the binding constraint on how hard the facility can be driven and how often. Final optics are consumable. That is the cost structure Xcimer is attacking.

Its answer is to do the hard part in gas. Phoenix, the prototype system the company began operating in June, is a krypton fluoride excimer laser that uses stimulated Brillouin scattering to compress a pulse roughly a microsecond long down to nanosecond timescales — and the SBS cell doing that compression is 38 metres of gas. A gas optic has no surface to damage in the way a coated substrate does; the medium flows, and what would be a replacement schedule for a solid-state component becomes a gas-handling problem instead. Xcimer puts over a kilojoule through Phoenix today, calls it the largest privately owned laser system in the world, and houses it in a 74,000-square-foot Denver facility. The company's commercial design targets megajoule-scale energy through two beamlines rather than hundreds.

Figure 1. The pulse compression cell in Xcimer's Phoenix laser system, photographed December 2024. Credit: Xcimer Energy.

Whether two beamlines at megajoule scale is easier than 192 at the same total is not settled, and nothing in the October announcement settles it. The energy per beamline rises by the same factor the count falls, and with it every aperture, every window and every pointing tolerance in the chain. What is not in dispute is the direction of the cost argument: fewer, larger, cheaper-per-joule, with the expensive consumable moved from a polished surface into a flowing medium.

The more consequential commercial event of Xcimer's autumn was not the DOE award. On 10 September RTX and Xcimer announced that RTX Ventures had taken a minority stake in the company and that the two had signed a teaming agreement to explore putting Xcimer's multi-kilojoule-class pulsed lasers into RTX defence systems, in containerised form, aimed at airborne threats and at longer ranges than current directed-energy systems reach. The investment amount was not disclosed. RTX reported 2025 sales above $88 billion.

Figure 2. Laser beam lines at Xcimer's Phoenix facility in Denver, Colorado. Credit: Xcimer Energy.

That matters for the same reason it has mattered in every other laser market. A fusion plant is a 2035 revenue line at the earliest; an air-defence programme is a procurement cycle. If the pulsed-power modules, the amplifier chambers and the gas handling can be sold into directed energy while the fusion physics is still being argued, the manufacturing learning curve gets paid for by someone else. The pattern should be familiar to anyone who watched fibre lasers reach industrial scale on the back of defence and then undercut everything in cutting and welding.

It also carries the usual risk. Defence programmes shape hardware around their own requirements — ruggedisation, duty cycle, transportability — and a laser optimised for a container on a flatbed is not necessarily the laser a power plant wants. Xcimer is betting the overlap is large enough, and the shared element it points to is the one that is genuinely common to both: a pulsed excimer architecture that fires repeatedly and reliably at low cost per joule.

Not the physics, and not the schedule. What changed on 6 October is that a third party with no equity in the outcome signed off on another tranche of technical milestones, which is the only external scoring system this sector currently has. For the optics industry the interesting signal is narrower and more durable: a well-funded company is building a megajoule-class laser on the premise that the cost of inertial fusion is dominated by optics, and that the way to win is to use fewer of them and to make the ones doing the violent work out of gas. If Xcimer is right, the suppliers who should be paying attention are not the ones polishing large apertures. They are the ones who can hold a window, a coating and an alignment tolerance at the two ends of a 38-metre cell.