Ampera unveils SMR design combining 3D printing and thorium fuel for AI data centers

Ampera, a South Korean energy startup, has unveiled a small modular reactor (SMR) design that integrates additive manufacturing for core component production and uses thorium-based fuel, targeting dedicated power supply for AI data centers. The company claims its 50 MWe SMR can be fabricated using metal 3D printing techniques, specifically laser powder bed fusion (LPBF) and directed energy deposition (DED), to reduce manufacturing lead times and enable complex internal geometries in heat exchangers and reactor vessel internals. Ampera has not disclosed a specific build timeline or regulatory filing date, but positions the design as a response to the growing energy demands of hyperscale AI computing clusters in South Korea and across Asia.

This announcement sits at the intersection of two emerging industrial trends: the energy sector's early-stage exploration of AM for nuclear components and the AI industry's desperate search for dedicated, carbon-free baseload power. Ampera's use of thorium - a more abundant and less proliferation-sensitive fuel than enriched uranium - adds a layer of differentiation, though thorium fuel cycle technology remains uncommercialized globally. The company's reliance on metal AM for pressure-bearing nuclear components places it in a high-qualification-burden environment similar to aerospace: any reactor part printed via LPBF or DED will require extensive regulatory certification from bodies like Korea's Nuclear Safety and Security Commission (NSSC) and potentially the U.S. NRC. Ampera is not alone - competitors like NuScale and X-energy have explored AM for prototyping, but no SMR developer has yet achieved regulatory approval for a production AM part in a reactor core. The AI data center angle is commercially pragmatic: it provides a captive, high-value customer willing to pay premium power prices, bypassing the grid interconnection delays that have plagued larger SMR projects.

From a practical standpoint, Ampera faces a multi-year qualification journey before any AM-produced SMR component enters service. The company must first demonstrate that its LPBF and DED processes can produce repeatable, inspectable, and radiation-tolerant parts - a challenge that has taken aerospace programs over a decade to solve. For the AM industry, this is a speculative but real frontier: if Ampera or any SMR developer successfully qualifies a printed nuclear component, it would open a new vertical for metal AM with extremely high barriers to entry and long revenue tails. For now, the announcement is a design-stage concept, not a production commitment, and should be evaluated as such.