- Argonne wants to build an AI supercomputing facility running 100,000 GPUs,
- NVIDIA agreed to be an anchor partner supplying GPUs, with Argonne open to more partners
- Feedback is due Aug. 21
Argonne National Laboratory has asked industry, academia and other organizations how they might help build and use a large-scale artificial intelligence supercomputing facility on the lab’s grounds. The Department of Energy laboratory posted a request for information on SAM.gov on Tuesday, with responses due Aug. 21.
The planned National Genesis Mission Supercomputing Center of Innovation would provide computing at the scale of 100,000 graphics processing units, which Argonne says would make it the world’s largest AI computing resource dedicated to science, engineering and medicine. NVIDIA has signed on as an anchor partner supplying GPUs and computing technologies, and the lab said it is open to additional partners.
Argonne’s push to scale AI computing reflects the kind of infrastructure investment reshaping federal civilian agencies, a theme the Potomac Officers Club’s 2026 FedCiv Summit will take up on Oct. 29. Register now!
How Does Argonne Want to Finance the AI Supercomputing Center?
Argonne is steering toward a private investment model, expecting much of the infrastructure to be financed through non-federal sources rather than taxpayer money. The lab said deploying the supercomputer requires a long-term partnership with one or more private sector integration or infrastructure partners, and that the first phase alone is estimated to require at least $6 billion in privately financed investment over its first 18 months. Organizations without large-scale capital to offer are still encouraged to respond regarding equipment, software, services, pilot projects or test beds.
What Is Argonne Asking Respondents to Address?
The RFI covers two focus areas: contributing computing and infrastructure capabilities, and using, co-funding or reserving capacity from the center. Argonne wants input on partnership and governance models, including how a consortium of universities and medical centers might be structured to provide researchers access to computing power beyond what individual institutions could obtain on their own.
