Purdue University

Researchers nationwide can now request access to the Rosen Center for Advanced Computing’s (RCAC) new national artificial intelligence (AI) resource, Anvil AI.

Purdue University's powerful national HPC-resource, the Anvil supercomputer, recently received an upgrade, thanks to support from the National Science Foundation’s (NSF) NAIRR Pilot Program. This hardware upgrade added a total of 84 Nvidia H100 SXM GPUs to the system. Along with the GPUs, Anvil AI also features 1 petabyte of object storage, which can host datasets needed by AI researchers. Now that the hardware has been successfully installed and tested, “Anvil AI” is available for allocation requests through the NAIRR Pilot.

When requesting allocations through NAIRR, researchers should choose “Purdue Anvil AI” over “Purdue Anvil GPU.” This ensures that the users will have access to the new H100 GPUs instead of the less powerful, A100 GPUs.

“The H100 GPU outperforms the current A100 GPU in Anvil by as much as nine times in computing speed,” says Rosen Center Chief Scientist Carol Song, principal investigator and project director for Anvil. “Many workloads, especially AI model training and inference, will run much faster, reducing the time-to-results for researchers.”

To learn more about the Anvil supercomputer’s architecture, please read our Anvil AI Hardware Announcement.

Researchers and educators can apply for access to NAIRR resources and view descriptions of the first cohort projects at https://nairrpilot.org/. Resource request submissions can be made following the process outlined in https://nairrpilot.org/opportunities/allocations. Submissions should select the “Purdue Anvil AI” resource and specify their resource requirements in terms of H100 GPU hours. Anyone with questions should contact anvil@purdue.edu.

Written by: Jonathan Poole, poole43@purdue.edu