US national labs seek new partnerships for AI supercomputers

The Department of Energy’s network of national laboratories is actively seeking fresh private-sector partners to build AI-optimized systems capable of powering the next wave of research.

The money and the machines

The 2025 AI Action Plan, combined with the “One Big Beautiful Bill Act,” allocates $150 million through 2026 specifically for AI initiatives at DOE laboratories.

Oak Ridge National Laboratory is set to receive two new AMD-accelerated AI supercomputers through a public-private partnership. The first deployment is expected soon, with a larger, more capable system slated for 2028.

Building the coordination layer

The DOE launched the National Laboratory AI Roundtable, a coordinating body involving eight national labs designed to synchronize AI standards and streamline deployment across the department’s research network.

The CM2US initiative aims to apply AI to securing the US critical minerals supply chain, using machine learning to accelerate domestic mineral discovery and extraction.

The security calculus

Analysts have flagged that integrating AI capabilities with National Nuclear Security Administration infrastructure could strengthen US defense capabilities while simultaneously creating new attack surfaces in expanded data-sharing environments between labs and private partners.

What this means for the tech landscape

AMD stands to benefit directly from the Oak Ridge partnership, which validates its data center and HPC accelerator strategy at a moment when the company is fighting to carve out market share against NVIDIA’s dominance in AI training workloads.