Meta lines up 6.6 GW of nuclear power to back AI data center growth
Original: Meta Announces Nuclear Energy Projects, Unlocking Up to 6.6 GW to Power American Leadership in AI Innovation View original →
Meta said on January 9, 2026 that it had signed nuclear energy agreements with Vistra, TerraPower, and Oklo as it prepares for a larger wave of AI infrastructure. The company said the projects could support up to 6.6 GW of new and existing clean power by 2035 and help supply the grids that support its operations, including the Prometheus supercluster in New Albany, Ohio.
The announcement matters because electricity availability is becoming a hard constraint for frontier AI. Instead of relying only on renewable credits or short-term power contracts, Meta is committing across several layers of the nuclear stack at once: extending the life of existing plants, backing uprates, and helping fund advanced reactor deployment. That gives the company a longer-duration path to firm power as model training and inference workloads keep scaling.
- Meta said its TerraPower agreement supports two new Natrium units capable of generating up to 690 MW, with delivery as early as 2032.
- The same agreement gives Meta rights to energy from up to six additional Natrium units capable of producing 2.1 GW by 2035, for a total of 2.8 GW of baseload generation and 1.2 GW of built-in storage.
- The Oklo project in Pike County, Ohio could add up to 1.2 GW of clean baseload power as early as 2030.
- Vistra will supply more than 2.1 GW from operating plants in Ohio and Pennsylvania, plus 433 MW of uprate capacity expected in the early 2030s.
Meta framed the deals as broader grid and industrial policy support, not just a private power reservation. The company said the projects should help preserve existing nuclear plants, accelerate next-generation reactor programs, strengthen the nuclear fuel supply chain, and create thousands of construction jobs along with hundreds of long-term operational roles.
The bigger signal is that AI infrastructure competition is now extending well beyond chips and models. Securing dependable electricity over a 10-year horizon is becoming part of the core operating strategy for hyperscalers and AI labs alike. By moving into nuclear at this scale, Meta is effectively betting that firm, dispatchable power will be a decisive input to the next generation of AI data centers.
Related Articles
Meta said on March 11, 2026 that it is developing and deploying four new generations of MTIA custom chips within the next two years. The company is positioning MTIA as a central part of its AI infrastructure strategy for ranking, recommendations, and GenAI inference workloads.
Meta says custom silicon is critical to scaling next-generation AI and has published a roadmap update for its MTIA family. The company says it accelerated development enough to release four generations in two years as model architectures keep changing faster than traditional chip cycles.
OpenAI and Pacific Northwest National Laboratory introduced DraftNEPABench, a benchmark for permitting work that uses coding agents on document-heavy NEPA tasks. In tests across 18 federal agencies, experts found the system could save 1 to 5 hours per subsection, or roughly 15% of drafting time.
Comments (0)
No comments yet. Be the first to comment!