Federal Project Grant Award Summary

The National Science Foundation (NSF), Division of Information and Intelligent Systems, awarded a CAREER grant of $308,232 to the University of Oregon on September 1, 2026, under the Computer and Information Science and Engineering program (CFDA 47.070). The project, titled "PROTRAIN: Enabling Efficient Large Language Model Training via Performance, Energy, and Reliability Co-Optimizations," will deliver a unified research and development framework addressing the substantial computational, energy, and reliability challenges associated with training large language models (LLMs). The primary deliverables include the development of a novel efficiency model that integrates performance, energy consumption, reliability, and model quality optimizations to enable data-driven, cost-aware decision-making for large-scale LLM training operations.

The five-year initiative (completion date August 31, 2031) will advance optimization methods in underexplored areas, particularly energy efficiency and reliability, while establishing quantitative models for understanding trade-offs among existing optimization strategies. The project also includes an integrated education and workforce development program designed to expand participation in advanced computing and the AI industry. Outcomes are anticipated to promote sustainable computing practices, strengthen U.S. national competitiveness in artificial intelligence, and support applications advancing economic growth, education, national security, and public services.