Project Grant 2514055

This federal Project Grant award, provided by the National Science Foundation's (NSF) Computer and Information Science and Engineering (CISE) program (CFDA 47.070), aims to enhance the performance and productivity of large-scale artificial intelligence (AI) applications through improvements to the Message Passing Interface (MPI), a widely used standard for coordinating work across parallel computing systems.

The $800,278 award, effective September 15, 2025 through August 31, 2029, supports a multi-university collaboration between Tennessee Technological University, the University of Tennessee, Knoxville, Stony Brook University, and the Illinois Institute of Technology. The project, titled "MPI4AI: Enhancing Performance and Productivity of AI Science through Next-Generation High Performance Communication Abstractions," will develop key improvements to the open-source Open MPI implementation, including native GPU communication support, enhanced collective communication operations tailored for AI algorithms, and optimized data movement. These advancements target performance bottlenecks in critical AI workflows such as neural architecture search, large language model inference, and large-scale data-parallel training. By contributing these enhancements to future MPI standards, the project aims to have long-term impact across academic research and industrial AI applications, lowering the cost of running large AI workloads and broadening access to scalable AI infrastructure.