Nvidia explores converting decommissioned German nuclear plants into AI data centers
Nvidia announced a preliminary partnership with German energy firm PreussenElektra GmbH to assess whether the utility’s decommissioned nuclear sites could be repurposed as high‑performance AI data centers. The two companies said the feasibility study will examine existing grid connections, land availability and industrial infrastructure to determine if the strict power and cooling requirements…
Key points
- Nvidia and PreussenElektra entered a preliminary partnership to study repurposing decommissioned nuclear sites for AI data centers
- The feasibility study will evaluate grid links, land and industrial infrastructure to meet AI factory requirements
- No investment, construction plans or use of nuclear power have been confirmed; the project is experimental
Guido Knott, chairman of PreussenElektra’s management board, said the former power‑plant locations have valuable infrastructure that could support new, future‑oriented uses. Nvidia’s senior regional director Markus Hacker added that reusing such sites could cut costs and shorten development timelines compared with building new facilities from scratch. The initiative remains experimental; no investment amounts, construction agreements or plans to use nuclear power for the data centers have been finalized.
Germany shut its last operational nuclear plant in April 2023 and is dismantling about 30 plants. PreussenElektra is responsible for decommissioning eight of them, which are the likely candidates for the assessment.
Nvidia eyes converting old nuclear plants into massive AI data centers
Tom's Hardware · 25 October 2026
Loading the full article…
This text was published by Tom's Hardware and written by Etiido Uko. It is reproduced here with attribution so you can read it in full; the rights remain with the publisher. Read it at the source ↗
The headline, key points and digest above were generated by Digest AI's editorial model from the linked sources. Automated summaries can contain errors: the sources are the record. Spotted a mistake? Tell us. Published by Martin K., who runs Digest AI and handles corrections.
More in Hardware & Compute
All →- Cerebras explains how to run 3T models on WSE-3T chips · 1 src
- Zagita tool doubles RTX 30/40 GPU compute, driving used prices up · 1 src
- Nvidia and SPIL plan 15x expansion of co-packaged optics capacity by mid-2027 · 1 src
- Investors watch AMD, Google, and Amazon challenge NVIDIA in AI chips · 2 src
- Google and Voltus expand virtual power plant deal for 1,070 megawatts in PJM grid · 1 src
Comments
via GitHub Discussions