Grouptuner: Efficient Group-Aware Compiler Auto-tuning

Bingyu Gao, Mengyu Yao, Ziming Wang, Dong Liu, Ding Li, Xiangqun Chen, Yao Guo · 2025

Read Grouptuner: Efficient Group-Aware Compiler Auto-tuning by Bingyu Gao, Mengyu Yao, Ziming Wang, Dong Liu, Ding Li, Xiangqun Chen, Yao Guo (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2505-08598.txt.

Description

Bingyu Gao and six co-authors speed up compiler tuning in a 2025 computer science paper. Compilers offer hundreds of switches that affect how fast a program runs, and standard bundles like -O3 rarely hit the peak for a given program. Auto-tuning searches the switch space, but the data it gathers is sparse and noisy.

Their method, GroupTuner, mutates whole groups of switches that appeared together in the best runs so far. It skips the shaky step of guessing which single switches matter most. In experiments it beat the -O3 default by about twelve percent on average. It also used about three-quarters of the time of random search and outperformed leading rival methods.

Catalog record

CREATORBingyu Gao, Mengyu Yao, Ziming Wang, Dong Liu, Ding Li, Xiangqun Chen, Yao Guo
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE54 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2505-08598

Browse the full library · About the Tower of Babel