一作OSDI 2026Oral· Operating Systems & Distributed Systems
Optimal Software Pipelining and Warp Specialization for Tensor Core GPUs
Rupanshu Soi · Rohan Yadav · Fredrik Kjolstad · Alex Aiken · Maryam Mehri Dehnavi · Michael Garland · Michael Bauer
摘要
GPU architectures have continued to grow in complexity, with recent incarnations introducing increasingly powerful fixed-function units for matrix multiplication and data movement to accompany highly parallel general-purpose cores. To fully leverage these machines, software must use sophisticated...