Presenter

Maria Garzaran is a senior principal engineer at Intel. She is the architect of oneCCL — the AI communication library that helps deep learning frameworks scale model training more efficiently. She also serves as the technical lead for the MPICH-based distributed communication library deployed on the Aurora exascale supercomputer, leading the design to enable Intel GPUs in MPICH. Maria is a well-recognized expert in parallel programming and architectures with over 70 global publications.