Ampera, a nuclear-technology startup, is developing a subcritical, solid-state thorium-based microreactor built around a 3D-printed silicon carbide core and pressure vessel using TRISO fuel particles, The Register reported July 3, 2026. The design targets 15 or 30 megawatts electric depending on configuration, aimed specifically at datacenters, defense applications and off-grid sites -- three customer segments where reliable, dense, on-site power is increasingly valuable and difficult to secure through conventional grid connections.
The company's central manufacturing bet is that 3D-printing the reactor core enables a genuinely factory-built, mass-produced approach to nuclear power generation, rather than the custom, one-off construction methods that have historically made nuclear projects slow and expensive to deploy at scale. Founder Brian Matthews framed the technology as setting the foundation for that factory-built model, describing it as offering 'a clear commercial path for new nuclear technology' rather than another speculative long-horizon research program.
Ampera's staged rollout plan separates the power-generation portion of the system, which the company says could be available as early as 2027, from the full nuclear module itself, targeted for around 2030 pending regulatory approval -- a structure that lets the company potentially generate revenue and prove out parts of its technology well before the more heavily regulated nuclear component clears approval.
The design's headline technical claim -- up to 30 years of operation without refueling -- would, if achieved, meaningfully change the maintenance and operating-cost calculus for datacenter operators weighing on-site nuclear power against continuous reliance on grid connections or natural gas generation, both of which carry their own supply and emissions tradeoffs covered elsewhere in this issue.
Ampera has already taken steps to secure its supply chain, establishing an Australian subsidiary in June 2026 specifically to secure thorium supplies and planning domestic TRISO fuel kernel production -- moves that suggest the company is thinking beyond core reactor design toward the full fuel and manufacturing pipeline required to actually deliver at scale.
For datacenter operators and AI infrastructure investors, Ampera's staged, factory-manufacturable approach is a notable entrant in the broader race to solve AI's power-demand crunch through advanced nuclear, alongside a growing field of other microreactor and small-modular-reactor startups pursuing similar customer bases. For energy investors, the 2027-to-2030 staged timeline gives a concrete, checkable set of near-term milestones for evaluating whether Ampera's manufacturing approach translates into actual deployed capacity on schedule.
What to watch: whether Ampera's power-generation component hits its 2027 availability target, how the company's regulatory approval process for the full nuclear module proceeds toward 2030, and whether the company signs its first announced datacenter or defense customer contracts.