Скачать или смотреть [LCTES'25] Grouptuner: Efficient Group-Aware Compiler Auto-tuning

Grouptuner: Efficient Group-Aware Compiler Auto-tuning (Video, LCTES 2025)
Bingyu Gao, Mengyu Yao, Ziming Wang, Dong Liu, Ding Li, Xiangqun Chen, and Yao Guo
(Peking University, China; Peking University, China; Peking University, China; ZTE Corporation, China; Peking University, China; Peking University, China; Peking University, China)

Abstract: Modern compilers typically provide hundreds of options to optimize program performance, but users often cannot fully leverage them due to the huge number of options. While standard optimization combinations (e.g., -O3) provide reasonable defaults, they often fail to deliver near-peak performance across diverse programs and architectures. To address this challenge, compiler auto-tuning techniques have emerged to automate the discovery of improved option combinations. Existing techniques typically focus on identifying critical options and prioritizing them during the search to improve efficiency. However, due to limited tuning iterations, the resulting data is often sparse and noisy, making it highly challenging to accurately identify critical options. As a result, these algorithms are prone to being trapped in local optima.
To address this limitation, we propose GroupTuner, a group-aware auto-tuning technique that directly applies localized mutation to coherent option groups based on historically best-performing combinations, thus avoiding explicitly identifying critical options. By forgoing the need to know precisely which options are most important, GroupTuner maximizes the use of existing performance data, ensuring more targeted exploration. Extensive experiments demonstrate that GroupTuner can efficiently discover competitive option combinations, achieving an average performance improvement of 12.39% over -O3 while requiring only 77.21% of the time compared to the random search algorithm, significantly outperforming state-of-the-art methods.

Article: https://doi.org/10.1145/3735452.3735530

Supplementary archive: https://doi.org/10.5281/zenodo.15348539 (Badges: Artifacts Available, Artifacts Evaluated — Reusable, Results Reproduced)

ORCID: https://orcid.org/0009-0005-3491-1384, https://orcid.org/0009-0005-8220-3470, https://orcid.org/0009-0008-1794-213X, https://orcid.org/0009-0009-3181-5994, https://orcid.org/0000-0001-7558-9137, https://orcid.org/0000-0002-7366-5906, https://orcid.org/0000-0001-5064-5286

Video Tags: Compiler Optimization, Compiler Auto-tuning, Compiler Options, pldiws25lctesmain-p68-p, doi:10.1145/3735452.3735530, doi:10.5281/zenodo.15348539, orcid:0009-0005-3491-1384, orcid:0009-0005-8220-3470, orcid:0009-0008-1794-213X, orcid:0009-0009-3181-5994, orcid:0000-0001-7558-9137, orcid:0000-0002-7366-5906, orcid:0000-0001-5064-5286, Artifacts Available, Artifacts Evaluated — Reusable, Results Reproduced

Presentation at the LCTES 2025 conference, June 16–17, 2025, https://pldi25.sigplan.org/home/LCTES...
Sponsored by ACM SIGPLAN, ACM SIGBED,