On July 20, 2026, the US Department of Energy’s National Nuclear Security Administration selected contractor Amentum to enter negotiations for a phased lease to build a 1-gigawatt AI data center at the Savannah River Site in South Carolina. The project pairs the facility with about 2 gigawatts of on-site natural gas and future nuclear generation to power advanced AI workloads without shifting costs to existing ratepayers.
This article aggregates reporting from 1 news source. The TL;DR is AI-generated from original reporting. Race to AGI's analysis provides editorial context on implications for AGI development.
The Savannah River project is a bellwether for how governments plan to provision industrial-scale AI compute. A 1GW data center coupled with 2GW of dedicated generation is effectively an AI power plant, explicitly framed by the Trump administration as part of a national strategy to ‘power the global AI race.’ Situating that infrastructure on federal nuclear land under NNSA oversight marries AI capacity with national security concerns, signaling that high-end compute is now treated as strategic infrastructure alongside shipyards and missile silos.
This kind of vertically integrated AI campus—compute, power, land, and federal backing in one package—could become a template for future deployments in the US and allied countries. It also highlights an emerging divide: jurisdictions willing to rezone and rewire their grids for AI will attract the most advanced models and workloads. For the AGI race, that means the bottleneck shifts less from “can we find power and land?” to “who is allowed to plug into these gigawatt-scale clusters, and under what rules?” Access, governance, and export controls around such facilities may matter as much as FLOP counts.


