- A Gigawatt Nuclear Energy Data Center could soon open in France
- “Power is the first consideration” for AI
- Models need more energy for training than ever
French President Emmanuel Macron announced 109 billion euros in private investments to improve French AI infrastructure.
The investment key is a nuclear energy gigawatt, which will be intended for computers and high performance data centers necessary to supply emerging AI tools.
Announced within the framework of the AI action of the AI action in Paris, the power will be added to the already large nuclear infrastructure of France, which consists of 57 reactors on 18 different factories.
France will use nuclear energy to supply AI
About a third of the country’s entire energy consumption comes from nuclear energy, and its infrastructure is so extensive that it exported excess energy to other countries last year. The use of nuclear energy to supply its AI data centers seems to be natural progression.
Macron added: “Plug, baby, plug” – a reference to the “exercises, exercises” of American president Donald Trump, plan to extend oil drilling operations through the states.
Josh Parker, principal director of the sustainability of NVIDIA companies, said at the summit: “Food is the first consideration to have access to AI and bring our online systems” (via Wsj).
The first wave of the project will consist of 250 nuclear energy megawatts estimated at AI by the end of 2026. Fluidstack, the company behind the project, aims to start working in the third quarter of 2024. To obtain most of its summit – NVIDIA performancecips – currently the second most precious company in the world with an assessment of 3.27 billions of dollars.
Energy requests only increase. Separated Wsj The report reveals that the main models of AI today used 30 megawatts of energy for training. This could reach five gigawatts at the end of the decade.
The latest investments add France to the breed of countries and companies that submit large sources of nuclear energy and other energy centers to fuel future data centers.