Understanding Performance Problems in CUDA Programs
Proceedings of the ACM on Software Engineering · 30.06.2026 · 3 (FSE) · 344-365
DOI: 10.1145/3797143 ↗Kurz gefasst
Fachartikel in der Zeitschrift Proceedings of the ACM on Software Engineering, online erschienen am 30.06.2026. Verfasst von Yuyang Bi und 6 weiteren Autor:innen. Die Arbeit nennt 53 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
With the wide adoption of GPUs, CUDA programming has become essential for leveraging GPU parallelism. However, its complex programming model poses challenges in performance optimization. Consequently, CUDA programs often suffer from performance problems. In that sense, it is crucial to understand the performance problems specific to CUDA programming. Unfortunately, no systematic study has been conducted in literature. To bridge this gap, we conduct the first systematic study to 1) characterize the symptoms and root causes of 216 performance problems collected from 55 StackOverflow posts and 122 NVIDIA forum posts, and 2) measure the speedup of fixing performance problems, and assess the capability of existing performance analysis methods in identifying performance problems, using a dataset of 69 reproduced performance problems. Our findings provide practical guidance for developers, and opportunities for researchers to advance performance analysis.
Zitieren
APA
Bi, Y., Cao, J., Lu, Y., Chen, B., Gan, T., Wang, D., & Peng, X. (2026). Understanding Performance Problems in CUDA Programs. Proceedings of the ACM on Software Engineering, 3(FSE), 344-365. https://doi.org/10.1145/3797143
Vancouver
Bi Y, Cao J, Lu Y, Chen B, Gan T, Wang D, et al. Understanding Performance Problems in CUDA Programs. Proceedings of the ACM on Software Engineering. 2026;3(FSE):344-365. doi:10.1145/3797143
BibTeX
@article{Bi2026,
author = {Bi, Yuyang and Cao, Junming and Lu, You and Chen, Bihuan and Gan, Tianwei and Wang, Dingji and Peng, Xin},
title = {Understanding Performance Problems in CUDA Programs},
journal = {Proceedings of the ACM on Software Engineering},
year = {2026},
volume = {3},
number = {FSE},
pages = {344--365},
doi = {10.1145/3797143}
}
EUVIMED