Nvidia, Microsoft Target Nuclear Energy to Power AI Data Centers

Microsoft and Nvidia Partner to Accelerate Nuclear Design with AI

Microsoft and Nvidia announced a collaboration on March 24 to apply artificial intelligence to the nuclear energy sector. The partnership will focus on developing AI-powered tools to streamline the complex design and licensing processes for new nuclear power plants. By automating and optimizing these stages—historically a major bottleneck for the industry—the tech giants aim to significantly shorten development timelines and reduce costs. This initiative opens a new market for Microsoft's and Nvidia's AI platforms, positioning their technology as a critical enabler for the expansion of carbon-free energy infrastructure needed to power the digital economy.

AtkinsRéalis Deploys Nvidia's Omniverse for Nuclear-Powered AI Factories

The collaboration is part of a wider convergence between AI and nuclear energy. Engineering firm AtkinsRéalis, which holds the license for CANDU nuclear technology, is separately working with Nvidia to plan and deliver "AI-ready" facilities. AtkinsRéalis is leveraging Nvidia's Omniverse platform to create digital twins of integrated infrastructure systems, allowing for the complete design and optimization of nuclear-powered data centers before any physical construction begins. The firm noted that "nuclear power is emerging as a leading low-carbon baseload for AI factories," highlighting the industry's view of nuclear as an essential power source for dedicated AI computing workloads. This approach uses AI to build the very infrastructure that AI itself requires.

AI Demand Drives Push to Quadruple US Nuclear Capacity to 400 GW

These corporate initiatives are supported by a major government push to expand nuclear power. The U.S. government has set a target to increase domestic nuclear capacity from approximately 100 GW today to 400 GW by 2050, responding directly to the surging electricity demand from AI and data centers. Tech leaders including Microsoft, Google, Amazon, and Oracle have all made public commitments to using nuclear energy to power their operations. The Idaho National Laboratory (INL) is already working with these companies to develop AI tools that have shown the potential to more than double the speed of reactor development, creating a feedback loop where AI accelerates nuclear deployment to meet its own power needs.