The Idaho National Laboratory and NVIDIA have partnered to apply artificial intelligence to nuclear reactor development, aiming to cut build timelines in half and reduce operating costs by more than 50% as the United States races to meet rising electricity demand. John Josephakis, the global vice president of Sales and Business Development for HPC/Supercomputing at NVIDIA, said:
NVIDIA is honored to collaborate with the U.S. government to apply AI and accelerated computing to advance nuclear energy, while reducing energy costs for Americans.
The collaboration is part of the Department of Energy Genesis Mission, a national effort focused on building a large scale scientific computing platform to accelerate discovery, strengthen national security, and advance energy innovation. Within that framework, the nuclear effort is codenamed Prometheus and targets faster, safer, and cheaper deployment of advanced reactors.
“This is the moment to decisively advance AI-accelerated nuclear energy deployment, increasing America’s energy affordability while also catalyzing the development of Artificial Intelligence in the United States,” said Rian Bahran, Deputy Assistant Secretary of Energy for Nuclear Reactors. “This public-private partnership presents a targeted approach to AI-acceleration that goes beyond incremental ‘uplift’ improvements. It has the potential to transform the paradigm for how we deploy nuclear energy in addition to how we advance R&D and discovery.”
Prometheus will rely on AI powered digital twins, generative models, and agent based workflows to simulate reactor systems before construction begins. Engineers will train these digital twins using decades of nuclear data, laboratory research, and operational records from INL, enabling them to validate designs, licensing pathways, and operational behavior in virtual environments instead of relying on slow physical iteration.
John Wagner, director of Idaho National Laboratory, said the partnership could fundamentally change how quickly nuclear energy comes online by automating engineering intensive processes while keeping human oversight in critical decisions. Officials say the approach could at least double reactor deployment speed while improving safety and cost predictability.
NVIDIA is contributing its GPU accelerated computing platforms and AI infrastructure, including the optimization of nuclear simulation codes such as MOOSE, BISON, Griffin, and Pronghorn to run at scale. The initiative will depend on the leadership-class supercomputers from the Department of Energy to train large-scale AI models and run reactor simulations. The facilities at INL, such as the Neutron Radiography Reactor and the Microreactor Applications Research Validation and Evaluation (MARVEL) project, will supply real-world validation data for digital twin models. However, MARVEL is not yet operational.
Officials say the long term goal is regulatory modernization and broader industry adoption, with future participation expected from reactor developers, utilities, and investors.
If successful, the Prometheus effort could reshape how nuclear power is built in the United States at a moment when reliable, carbon free energy is becoming a strategic priority.