Scientific models are today limited by compute resources\, forcing approximations driven by feasibility rather than theory. They cons equently miss important physical processes and decision-relevant regional details. Advances in AI-driven supercomputing — specialized tensor accel erators\, AI compiler stacks\, and novel distributed systems — offer unp recedented computational power. Yet\, scientific applications such as ocea n models\, often written in Fortran\, C++\, or Julia and built for traditi onal HPC\, remain largely incompatible with these technologies. This gap h ampers performance portability and isolates scientific computing from rapi d cloud-based innovation for AI workloads. In this talk\, we bridge that g ap by transpiling existing programs using the MLIR compiler infrastructure . This process enables advanced optimizations\, deployment on AI hardware\ , and automatic differentiation. In particular\, we demonstrate execution of a state of the art Julia-based ocean model (Oceananigans)\, with >\;2 77 custom single-node CUDA kernels on thousands of distributed GPUs and Go ogle TPUs. Our results demonstrate that cloud-based hardware and software designed for AI workloads can significantly accelerate simulations\, openi ng a path for scientific programs to benefit from cutting-edge computation al advances.
Bio William Moses is an Assistant Prof essor at the University of Illinois in the Computer Science and Electrical and Computer Engineering departments. He received a Ph.D. in Computer Sci ence from MIT\, where he also received his M.Eng in electrical engineering and computer science (EECS) and B.S. in EECS and physics. William's resea rch involves creating compilers and program representations that enable pe rformance and use-case portability\, thus enabling non-experts to leverage the latest in high-performance computing and ML. He is known as the lead developer of Enzyme\, a tool for LLVM/MLIR capable of differentiating code in a variety of languages\; Polygeist\, a polyhedral compiler and C++ fro ntend for MLIR\; and Reactant\, a tool for enabling existing scientific co de to run on distributed ML accelerators. He has also worked on the Tensor Comprehensions framework for synthesizing high-performance GPU kernels of ML code\, the Tapir compiler for parallel programs\, and compilers that u se machine learning to better optimize. He is a recipient of the SIAM SC E arly Career Prize\, the SIGHPC Doctoral Dissertation Award\, a DOE Computa tional Science Graduate Fellowship and the Karl Taylor Compton Prize\, MIT 's highest student award.
LOCATION:William Gates Building\, LT2 END:VEVENT END:VCALENDAR