Student: Mohammad Zaeed
Advisor: Dr. Tanzima Islam
Title: Analysis and Visualization of Important Performance Counters To Enhance Interpretability of Autotuner Output
Date/Time: 10 AM to 11 AM, 30th November, Wednesday, 2022.
ZOOM MEETING I.D. : https://txstate.zoom.us/j/98698827193
Abstract:
Autotuning is a widely used method for guiding developers of large-scale applications to achieve high performance. However, autotuners typically employ black-box optimizations to recommend parameter settings at the cost of users missing the opportunity to identify performance bottlenecks. Performance analysis fills that gap and identifies problems and optimization opportunities that can result in better runtime and utilization of hardware resources. This work combines the best of both worlds by integrating a systematic performance analysis and visualization approach into a publicly available autotuning framework, GPTune, to suggest to users (1) how to change tuning parameters to reduce application runtime, and (2) which hard performance counters correlate most with application runtime.
Deadline: Dec. 27, 2022, midnight