National High-Performance Computing – ACCESS

ACCESS (Advanced Cyberinfrastructure Coordination Ecosystem: Services & Support), formerly XSEDE, is a National Science Foundation-funded, nationwide collection of supercomputing systems available to researchers through merit-based allocations. The resources are free, but the application process can be competitive depending on the compute resources needed.

Columbia researchers can use a few types of allocations in order of ease of acquisition and amount of resources available.

Columbia Discover Allocation

• Use Columbia's Discover allocation (which Columbia ACCESS representatives from the Research Computing Services Team can readily approve).
• Recommended for small-scale testing and benchmarking.

Explore Allocation

• Explore requests require minimal documentation: CV and an abstract of the project.
• This allocation type is best for resource evaluation, graduate student projects, small classes and training events, benchmarking, code development and porting, and similar small-scale uses.

Accelerate Allocation

• Accelerate ACCESS requests require the following documentation: a project description, a three to five page project description and the PI's CV. A panel will review your request for approval.
• Accelerate is recommended for experienced users with mid-scale resource needs, consolidating multi-grant programs, collaborative projects, and gateways with growing communities.
• You would need to submit a Maximize access allocation request for resources greater than this.

Maximize Allocation

• Maximize requests are large-scale projects that require massive computational resources.
• Requests require the following documentation: a project description, a proposal no more than ten pages long, and the PI’s CV. A panel will review your request for approval. The requirements may evolve over time to ensure that the reviewers have adequate information to assess submissions and ensure efficient and effective use of the allocated resources.

Browse NSF ACCESS Resources

NAIRR 

The National Artificial Intelligence Research Resource (NAIRR) Pilot aims to connect U.S. researchers and educators to computational, data, and training resources needed to advance AI research and research that employs AI. Applications for NAIRR pilot resources are now open and available to the community.

Resources include:

• advanced computing systems
• cloud computing platforms
• access to models, software and privacy enhancing technology tools and collaborations to train models
• and education platforms

Many of NAIRR's resources are part of the ACCESS program and Campus Champion representatives at CUIT can help provide guidance to researchers who are interested in getting access to NAIRR resources.

Focus Areas:

This call for allocation proposals encompasses projects covering any of the cross-cutting or domain focus areas for the NAIRR Pilot.

Cross-cutting focus areas:

• Advancing AI methods that enable scientific discovery.
• Creating open-source foundation models for specific applications.
• Using large-scale models to explore complex datasets interactively.
• Advancing approaches for integrating simulations and AI.
• Using experimental data from sensors, detectors, or other edge instruments.
• Empowering use of scientific data by other stakeholders and the public.
• Training and educating a cohort of scholars in AI technologies and their responsible use.
• Probe key AI challenges using sensitive data including privacy-preserving methods.

Domain specific focus areas:

• Accelerating societally-relevant research on AI safety, reliability, security, and privacy.
• Empowering advances in cancer treatment and individual health outcomes.
• Supporting resilience and optimization of agricultural, water, and grid infrastructure.
• Improving design, control, and quality of advanced manufacturing systems.
• Addressing earth, environmental, and climate challenges via integration of diverse data and models.

Other projects that align with the broader objectives of the NAIRR Pilot, as well as projects in other areas of AI research and applications, may secondarily be considered for allocation.

How to Apply:

Researchers and educators can apply for access to these resources and view descriptions of the first cohort of awarded projects at nairrpilot.org. For those with a project in mind, you can apply to the following open calls:

• NAIRR Pilot Resource Request to Advance AI Research
• Classroom/Educator Resources Call
• For assistance, watch the Requesting Resources via the National AI Research Resource (NAIRR) Pilot webinar from March 31, 2025.

Project Expectations:

All persons submitting proposals should note the following parameters and expectations:

• Because of the strong public component of the NAIRR, all project results must be open and publishable. Teams who receive NAIRR Pilot access are expected to publish their results in the open scientific literature or otherwise disseminate publically, and to make all project products publically available, to the extent possible.
• Project PIs are expected to provide brief (2–3 paragraphs) project updates 1 month and 6 months into the project, and a final report (three pages) upon project completion.
• NAIRR Pilot projects are intended to be completed within twelve (12) months. Projects with longer-term timelines should contact [email protected] to be referred to existing proposal and allocation opportunities.
• To see full details and additional expectations, please check out NAIRR Pilot - Expectations of Project