Tech giants Microsoft and NVIDIA are collaborating on an artificial intelligence project designed to accelerate the development of nuclear energy – in order to feed the growing energy needs of AI. The project aims to develop an “ecosystem of AI-powered digital engineering tools” that will be used to shorten the considerable timelines of nuclear power plants and bring them online a lot more quickly as the rate of energy demand growth continues to skyrocket around the world.
The nuclear power sector in the United States is beset with bottlenecks ranging from complicated and costly bespoke design and engineering processes to lengthy regulatory processes characterized by miles and miles of red tape. The most recent nuclear power plant to come online in the United States, Georgia’s Plant Vogtle, showcased exactly how devastating and extensive these delays can be. When the plant was finally completed in April of 2024, it had taken 15 years and $35 billion to complete, making it the most expensive infrastructure project of any kind in United States history.
“The project has been such a bloated disaster that many pundits think it could be make-or-break for the wholesale future of the United States nuclear sector,” I wrote about Vogtle back in 2024. “But there are two ways to interpret the cautionary tale presented by Vogtle: either you think that the lesson is not to build new reactors, or the lesson is to build nuclear reactors better.”
Big Tech is clearly choosing the latter option. Nuclear power has gained favor in Silicon Valley as a potential solution to its runaway energy demand, which has been pushed into overdrive by rapid AI integration. As the public and the government ratchet up pressure for Big Tech to solve its energy problem – the burden of which currently rests on consumers, whether they support or benefit from AI or not – many tech bigwigs are starting to pour investment dollars into the nuclear sector.
Enter the “Digital Age for Nuclear.”
Microsoft and NVIDIA are throwing their enormous weight behind solving the major hurdles preventing a new nuclear era in the United States. They think that digitizing the analog processes upholding the sector will be a game-changer for allowing more efficient growth. This would be critical for allowing nuclear power generation capacity growth to come anywhere close to keeping pace with energy demand growth from data centers.
The new ‘ecosystem’ being rolled out by the two companies “provides end-to-end tools that combine AI and digital twins for creating faster iterative design and engineering solutions,” Interesting Engineering described in a recent report. “Licensing and permitting is handled by Generative AI for document drafting and gap analysis.”
Related: 5 Stocks to Buy Now That The Strait of Hormuz is Closed
More sophisticated modelling capabilities will also ease the process of designing new reactors. “While traditional 3D models only map physical space, 4D (time scheduling) and 5D (cost tracking) simulations can virtually construct the plant before shovels hit the dirt,” states a Microsoft press release related to the project.
These benefits are not just theoretical – Microsoft says that they are already seeing efficiency gains thanks to the collaborative initiative. The toolkit is already being rolled out in smaller spaces like Aalo Atomics and Idaho National Lab. And the results have been astonishing. Aalo has reported a 92% reduction permitting timelines, resulting in an estimated annual savings of $80 million.
“Two things matter most: enterprise-scale complexity and mission-critical reliability,” Yasir Arafat, CTO of Aalo Atomics, was quoted by Interesting Engineering. “We’re deploying something complex at a scale only a company like Microsoft really understands.”
In addition to streamlining the development and deployment of conventional nuclear reactors, Big Tech is also heavily invested in unlocking commercial nuclear fusion, which many proponents view as a sort of silver-bullet solution to creating huge amounts of energy without compromising on climate goals or producing hazardous nuclear waste. And, once again, they’re counting on AI to crack the code.
By Haley Zaremba for Oilprice.com