OpenAI and Microsoft are bringing their latest AI technology to US national laboratories, with plans to install an o1 or similar o-series model on NVIDIA's Venado supercomputer at Los Alamos National Laboratory (LANL).
The supercomputer will serve as a shared resource for researchers across three national labs: Los Alamos, Lawrence Livermore and Sandia. About 15,000 scientists at these facilities plan to use the AI models for projects ranging from medical research to cybersecurity and protecting critical infrastructure.
The initiative builds on OpenAI and Microsoft's existing partnership, with Microsoft continuing its role as the preferred provider of AI computing infrastructure.
The labs aim to apply these AI models to advance energy research and deepen their understanding of fundamental mathematical and physical relationships.
Nuclear safety takes center stage
A key focus of the collaboration is the labs' nuclear safety program, which works to reduce the risks of nuclear war while securing nuclear materials and weapons worldwide.
Given the sensitive nature of this work, OpenAI plans to carefully review each use of the AI models in nuclear-related applications. OpenAI researchers with appropriate security clearance will provide guidance on these projects. The company sees this partnership as a step toward using AI to advance science, enhance national security, and support government initiatives.
This isn't OpenAI's first collaboration with Los Alamos - the organizations previously worked together to assess how advanced AI models might affect the development of bioweapons, including artificially generated pandemics.
