Nvidia to build supercomputer for federal AI research

Nvidia is building a $20 million supercomputer with MITRE, a federally funded, not-for-profit research organization that has been a technological clearinghouse for the government since the 1950s.

The move comes in support of President Biden’s 2023 executive order on encouraging advances in AI. The U.S. government will use an Nvidia DGX SuperPOD to provide researchers and developers access to much more computing power than they have had in the past to produce generative AI advances in areas such as climate science, healthcare and cybersecurity.

The DGX SuperPOD powering the sandbox is capable of an exaFLOP of AI performance to enable researchers and developers to train and deploy custom large language models (LLMs) and other AI solutions at scale. Nvidia notes in the announcement that today, few federal agencies have adequate access to large-scale computing infrastructure, which limits what public sector use of AI tools like LLMs.

“There’s huge opportunities for AI to make government more efficient,” said Charles Clancy, senior vice president of MITRE said in a statement. “Government is inefficient, it’s bureaucratic, it takes forever to get stuff done. … That’s the grand vision, is how do we do everything from making Medicare sustainable to filing your taxes easier?”

MITRE was spun out of a MIT lab in 1958 and has been a supplier of surveillance, communications and cybersecurity technologies to the Pentagon and U.S. intelligence agencies. It operates numerous facilities around the country including the Fermi National Accelerator Lab.

The DGX SuperPOD will support MITRE’s Federal AI Sandbox, a platform to improve experimentation with next-generation, AI-enabled applications across federal government agencies. The sandbox launches in late 2024.