This call is open to meritorious proposals by US-based researchers and educators from US-based institutions including academic institutions, non-profits, federal agencies or federally funded R&D centers; state; local; or tribal agencies; startups and small businesses with federal grants. However, the individual resources available via this call may have differing eligibility rules. Assignment of supported proposals will be guided by these constraints.

References need not count against the three-page limit. The main document must still be no more than three pages and address all the proposal criteria. You may attach a separate References document (up to 5 pages).

Requestors are asked to justify the amount of requested resources based on the proposal’s nature, scientific significance, and computational needs. If your proposal is deemed meritorious and the estimated resource need exceeds availability of the selected resource, the project will be moved to a resource or resources that can accommodate the need to the extent possible based on availability of resources.

The "Choose for me" option will not affect how your proposal is considered. If you select that option, you should still provide details that help the reviewers understand the kinds of resources that would be appropriate for your project. For example, if you know you need GPUs, the estimated number of GPU-hours and the types of GPUs on which your codes have already been run would be useful information.

Typically, requests submitted by the 15th of the month will be reviewed and have their outcome decided by the end of the following month.

Yes, focus areas can be found at https://nairrpilot.org/opportunities/allocations

There are no defined upper limits on the amount of computational resources you can request either on a specific resource, or in aggregate if you request multiple resources.

It is recommended that you select one of the available resources that is suitable for your work, unless your project has a specific need to run at multiple sites—for example, in the case of a distributed learning framework.

Descriptions of several partner-contributed resources were given at a recent NSF-hosted Partner Showcase webinar, the recording of which can be found at https://new.nsf.gov/events/nairr-pilot-partner-showcase/2024-05-21. Future information sessions on available resource will be held over the following months.

The NAIRR pilot currently provides infrastructure resources for research engagements, and does not provide funding for the research work itself. As such, applicants are expected to have existing funding sources to support the research activities.

Should a given project consist of multi-disciplinary efforts involving a team of AI researchers, domain scientists, and/or cyberinfrastructure specialists who are undertaking specific research challenges that innovatively use, integrate with, or rely on one or more NAIRR Pilot requestable resources, they may wish to consider responding to the recent NSF Dear Colleague Letter for NAIRR Pilot Demonstration Projects.

Privacy and security capabilities are indeed implemented on a per-resource basis. As such privacy and security capabilities may differ between different resource providers.

You may reach out to NSF staff at NAIRR_Pilot@nsf.gov should you wish to become a contributing resource partner with the NAIRR Pilot.