NSF AI grant $10M: "Expanse 2," a national shared supercomputer serving the long tail of science and AI (UC San Diego)
NSF awarded about $10 million to the San Diego Supercomputer Center (SDSC) at UC San Diego to deploy "Expanse 2," a next-generation supercomputer. With the latest processors and accelerators and a high-speed Ethernet network, it is offered as a national resource that brings data-intensive computing to a broad base of researchers — the "long tail" of science and AI.
Grant overview (primary data)
- Award amount$10,000,000
- RecipientUniversity of California-San Diego(CA)
- ProgramInnovative HPC
- Period2026-06-01 〜 2028-05-31
- FunderU.S. National Science Foundation (NSF) / NSF
Key points
- Deploys "Expanse 2," a multi-petaflop/s shared supercomputer (successor to Expanse)
- Latest processors and accelerators plus a high-speed Ethernet network for data-intensive work
- Supports: chip design automation, AI drug discovery, digital-agriculture AI, fusion turbulence, open inference models
- Design philosophy of serving the "long tail" — not just giant projects
- Part of NSF Innovative HPC; ~$10M; UC San Diego (SDSC); 2026–2028
The U.S. National Science Foundation (NSF) awarded about $10 million ($10,000,000) for "Expanse 2," a next-generation supercomputer at the San Diego Supercomputer Center (SDSC), UC San Diego (NSF Award 2614012; program: Innovative HPC; June 2026–May 2028).
Per the abstract, Expanse 2 is a multi-petaflop/s system that combines the latest-generation processors and accelerators (e.g., GPUs) with an advanced Ethernet-based interconnect to support data-intensive workloads. It responds to national R&D priorities with the stated goal of enhancing national competitiveness. The abstract lists supported research areas including electronic chip design automation (EDA), AI-accelerated drug discovery, AI applied to digital agriculture, the study of instabilities and turbulence in fusion, and open inference models.
As the name suggests, this is the successor to the existing shared supercomputer "Expanse." The title's phrase — "continuing to serve the long tail of science and AI innovation" — captures the design philosophy: delivering computing not only to a few giant projects but to the many modest-scale yet numerous research efforts that make up the "long tail."
Why it matters: high-performance computing (HPC) is often assumed to be the preserve of a few very large projects, but Expanse 2 emphasizes the long tail — providing access to the many researchers and institutions that cannot field large machines on their own. As AI's compute demand surges, a nation maintaining shared supercomputing as a public good and opening it across fields — drug discovery, agriculture, chip design, fusion — helps close gaps in access to compute. Like other NSF awards this site covers (the National Data Platform, the cross-disciplinary data fabric MESA), it is an example of building national research infrastructure to accelerate science with AI.
Why it matters
A concrete example of a nation maintaining shared supercomputing as a public good and opening it across fields as AI compute demand surges. The lens of closing compute-access gaps is broadly useful for thinking about research-infrastructure policy and AI foundations.
FAQ
What is the "long tail of science"?
What is an accelerator?
Sources (primary)
Source: NSF Award Search (U.S. National Science Foundation, public domain). Amounts are the obligated amount. For privacy, we do not handle principal investigator names.
- NSF Award (original, official)
- NSF Award ID: 2614012