NSF Commits $100 Million to Regional AI Infrastructure Hubs

The U.S. National Science Foundation launched a $100 million program on August 4 to build state and regional AI infrastructure hubs, with NVIDIA, AMD, Intel, and Dell Technologies among the private-sector partners. The NSF State and Regional Artificial Intelligence Infrastructure Hubs program will fund up to 10 regional consortia that pool compute, data, and expertise for researchers, students, and educators who currently lack access to AI-scale computing resources. Public-private cost

2 min
NSF Commits $100 Million to Regional AI Infrastructure Hubs

The U.S. National Science Foundation launched a $100 million program on August 4 to build state and regional AI infrastructure hubs, with NVIDIA, AMD, Intel, and Dell Technologies among the private-sector partners.

The NSF State and Regional Artificial Intelligence Infrastructure Hubs program will fund up to 10 regional consortia that pool compute, data, and expertise for researchers, students, and educators who currently lack access to AI-scale computing resources.

Public-private cost share

The program uses a cost-share model. State or multistate consortia, drawing contributions from universities, state and local governments, philanthropies, and industry, will build and operate the actual compute infrastructure. NSF's $100 million funds the connective tissue: consortium coordination, workforce development, and faculty training. One award will be made per state or region.

NSF AI Infrastructure Hubs framework diagram

"Artificial intelligence is transforming how we conduct research, accelerate scientific discovery and address complex challenges across disciplines," said Brian Stone, performing the duties of the NSF director.

NVIDIA said it will contribute training resources, educator enablement, applied learning content, technical guidance, and partner platforms to participating hubs. AMD, Intel, Dell Technologies, Hangar, and the Secunda Innovation Fund are also named as supporters.

Built on existing templates

The hub model traces back to a 2020 partnership between NVIDIA, cofounder Chris Malachowsky, and the University of Florida, which turned UF into what NVIDIA calls the country's first AI-integrated university. That initiative now includes more than 300 AI-focused faculty across all 16 colleges and has drawn over $511 million in AI research awards since 2017.

The program also extends the NSF-led National Artificial Intelligence Research Resource (NAIRR) pilot, which backed more than 700 research projects over two years. NSF encourages the new hubs to integrate with NAIRR, which would let regional consortia tap additional compute capacity and share datasets.

Part of the White House Genesis Mission

The hubs program supports the White House-led Genesis Mission, the national AI-for-science effort established by executive order in November 2025. It also responds to a July 2026 White House Office of Science and Technology Policy report calling for expanded R&D infrastructure.

"Regional partners who share in the benefits of discovery will pool their resources to unlock compute at a scale that no individual stakeholder, and no federal program, could achieve alone," said OSTP Director Michael Kratsios.

NSF is leaving the infrastructure design to the regions. Consortia can run on-premises systems, cloud computing, or a combination, designing resources around regional economic priorities. The solicitation number is NSF 26-513, with an initial cohort of up to 10 hubs.

Sources

Written by

More to read

  • Auxiliary-Loss-Free Load Balancing in Mixture-of-Experts: How Dynamic Bias Adjustments Eliminate Gradient Conflict and Routing Collapse

    Sparse Mixture-of-Experts (MoE) architectures decouple parameter count from per-token compute cost by activating only a small subset of feed-forward network (FFN) parameters for any given token. While dense transformers evaluate every parameter across all sequence positions, MoE models route tokens dynamically to specialized sub-networks, enabling parameter scaling to hundreds of billions or trillions of parameters at the inference and training cost of much smaller dense models. However, condit

    1 min
  • Confidential LLM Inference in Production: Hardware TEEs, GPU Enclaves, Attestation, and Serving Performance Trade-Offs

    Confidential LLM Inference in Production: Hardware TEEs, GPU Enclaves, Attestation, and Serving Performance Trade-Offs Deploying large language models in multi-tenant cloud environments introduces a fundamental security boundary problem. Standard transport encryption (TLS) secures prompts in transit, and encryption-at-rest protects checkpoints on disk, but model weights, prompt tokens, and key-value (KV) caches exist in plaintext within system memory during active inference. For organizations p

    1 min
  • Google DeepMind Outlines 15-Year Game AI Arc and EVE Online Research Sandbox

    Google DeepMind has detailed its 15-year trajectory of game-based artificial intelligence research, outlining how milestones from arcade reinforcement learning to modern multimodal models have culminated in an experimental research program inside the persistent virtual universe of EVE Online. The retrospective connects early breakthroughs in discrete, fully observable games to the frontier challenges currently facing autonomous systems: long-horizon planning, non-stationary multi-agent dynamics

    1 min