Idaho National Laboratory has joined forces with technology giant Nvidia in an ambitious collaboration aimed at dramatically accelerating the deployment of advanced nuclear reactors across the United States. The partnership, announced publicly, targets cutting reactor development timelines in half while reducing operational expenses by 50 percent—a significant step toward making nuclear energy more economically competitive and faster to deploy.
The initiative, called the Prometheus project, operates under the umbrella of the Department of Energy’s Genesis Mission, which published 26 distinct challenges and corresponding artificial intelligence solutions for advancing the nation’s energy infrastructure. INL and Nvidia’s collaboration specifically addresses the challenge of “Delivering Nuclear Energy that Is Faster, Safer, and Cheaper,” positioning the partnership to reshape how the nuclear industry approaches design, regulatory approval, manufacturing, construction, and operational management, as first reported by the American Nuclear Society.
Creating a Virtuous Cycle of Energy and Computing Power
The Prometheus framework creates what partners describe as a mutually reinforcing cycle: artificial intelligence accelerates nuclear reactor development and deployment, while the resulting nuclear energy supplies the massive amounts of reliable baseload electricity that advanced computing and artificial intelligence infrastructure demands. This relationship addresses a critical bottleneck in the expansion of data centers and computing facilities, which currently struggle with energy constraints in many regions.
INL Director John Wagner emphasized the transformative potential of the approach. “By leveraging AI to design, license, and operate reactors, we can fundamentally change the timeline for bringing advanced nuclear energy on line,” Wagner stated, according to reporting by the American Nuclear Society.
Nvidia’s global vice president of sales and business development, John Josephakis, highlighted the complementary strengths each partner brings to the effort. As reported by the American Nuclear Society, Josephakis said that “Combining INL’s decades of nuclear expertise with Nvidia AI infrastructure will put AI to work to design, license, and operate reactors faster, safer, and at lower cost.”
Expanding the Nuclear Innovation Ecosystem
INL and Nvidia have outlined plans to potentially broaden the partnership beyond their initial collaboration. The two organizations indicate they may invite nuclear reactor developers, utility companies, investors, and other national laboratories to join the Prometheus effort, creating a more expansive ecosystem for nuclear innovation. Such an expansion could accelerate knowledge-sharing and resource deployment across the industry.
The collaboration will focus intensively on several critical areas: accelerating AI-powered approaches to reactor design and licensing processes, advancing manufacturing and construction methodologies, improving operational efficiency, strengthening industry competitiveness, upgrading supercomputing capabilities, validating data accuracy, and optimizing computational code performance. These broad objectives signal an intent to touch nearly every phase of the nuclear development and deployment lifecycle.
Leveraging Idaho’s Nuclear Assets for Validation
A distinctive advantage INL brings to the partnership is access to unique nuclear facilities and decades of accumulated data. The laboratory will deploy its legacy nuclear datasets, on-site research reactors, and digital twin technology to validate AI-driven designs and simulations. Two facilities stand out as particularly valuable for this work: the Neutron Radiography Reactor (NRAD) and the Microreactor Applications Research Validation and Evaluation (MARVEL) facility. MARVEL, currently under development at the Idaho facility, is expected to become operational by late 2027, providing a cutting-edge testing ground for advanced reactor concepts.
The partnership will also accelerate several nuclear simulation codes on Nvidia’s GPU computing architectures. These codes—designated MOOSE, BISON, Griffin, and Pronghorn—represent critical tools for modeling nuclear reactor performance and behavior. By optimizing them to run on modern graphics processing units, the collaboration aims to dramatically reduce computation times for complex reactor simulations, enabling faster iteration and refinement of designs.
What Comes Next
The Prometheus project represents a significant convergence of Idaho’s long-standing role as the nation’s premier nuclear research hub and the accelerating computational demands of the artificial intelligence era. As the partnership develops, outcomes will likely influence not only Idaho’s nuclear ambitions and expansion plans at INL but also the broader national strategy for deploying advanced reactors to meet growing electricity demand. The success of INL and Nvidia’s collaboration could reshape how quickly new reactor designs move from conception through licensing to commercial operation—potentially positioning Idaho facilities as central to the nation’s energy and technology future.