KPerfIR: Towards an Open and Compiler-centric Ecosystem for GPU Kernel Performance Tooling on Modern AI Workloads

Yue Guan, Yuanwei Fang, Keren Zhou, Corbin Robeck, Manman Ren, Zhongkai Yu, Yufei Ding, Adnan Aziz
July, 2025 Systems
Cite DOI PDF arXiv

Abstract

KPerfIR proposes a compiler-centric ecosystem that unifies GPU kernel introspection, performance analysis, and optimization pipelines for state-of-the-art AI workloads.

Type
Conference paper
Publication
Proceedings of the 20th USENIX Symposium on Operating Systems Design and Implementation (OSDI)
GPU Performance Tooling Compilers