AI data services separate access from permanent storage
Retrospective: National services need a plan beyond computing power.
- Event / announcement date
- 2026-07-22
NSF announced $83 million for integrated data services on July 22. Its planned platforms connect scientific datasets, computing and AI models; UC San Diego’s National Data Platform targets reproducible workflows across distributed resources. [1]
Researchers need to find and reuse data across projects, beyond obtaining computing time. NSF describes infrastructure being built or expanded, rather than completed scientific results. [1]
An April 15 program solicitation makes the boundary concrete. IDSS supports national services and requires developed software to carry an identified open-source license, but excludes permanent storage and curation costs for the research data itself. It encourages partnerships or fees for those costs: national access still needs a durable data-custody plan. [2]
The source
- NSF: Data infrastructure awards ↗
Source date 2026-07-22 · reviewed October 29, 2026 · publication - NSF: Data services solicitation ↗
Source date 2026-04-15 · reviewed October 29, 2026 · publication
About this item: Retrospective report. No first-hand reporting or independent testing is claimed. Corrections policy.
