Overview
- Fervo Energy, NVIDIA and the Pacific Northwest National Laboratory said on Monday, June 22, 2026 that they will develop EGS‑Twin, a digital twin platform that combines live field data, physics‑based reservoir models and AI forecasting.
- PNNL will begin training scalable AI models immediately using Fervo’s proprietary Nevada and Utah site data on NVIDIA hardware and U.S. Department of Energy supercomputers, and the trained models will be added to NVIDIA’s Omniverse libraries.
- The partners say the platform is designed to deliver near real‑time insight into subsurface changes so operators can adjust injections, monitoring and well placement to optimize heat recovery and power output.
- Fervo’s stock rose about 8% in premarket trading after the announcement even though the company reported a large Q1 2026 loss and minimal revenue, highlighting investor interest in the partnership despite weak near‑term financials.
- If proven at scale, EGS‑Twin could reduce geological uncertainty and lower per‑well drilling costs, which would help make enhanced geothermal systems more financeable, but benefits will require continued data, validation and time before they are realized.