ISCAS OpenIR
an insightful program performance tuning chain for gpu computing
Jia Haipeng; Zhang Yunquan; Long Guoping; Yan Shengen
2012
Conference Name12th International Conference on Algorithms and Architectures for Parallel Processing, ICA3PP 2012
SourceLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Pages502-516
Conference DateSeptember 4, 2012 - September 7, 2012
Conference PlaceFukuoka, Japan
Indexed TypeEI
ISSN0302-9743
ISBN9783642330773
Department(1) Lab. of Parallel Software and Computational Science Institute of Software Chinese Academy of Sciences China; (2) College of Information Science and Engineering Ocean University of China China; (3) State Key Laboratory of Computing Science Chinese Academy of Sciences China; (4) Graduate University of Chinese Academy of Sciences China
English AbstractIt is challenging to optimize GPU kernels because this progress requires deep technical knowledge of the underlying hardware. Modern GPU architectures are becoming more and more diversified, which further exacerbates the already difficult problem of performance optimization. This paper presents an insightful performance tuning chain for GPUs. The goal is to help non-expert programmers with limited knowledge of GPU architectures implement high performance GPU kernels directly. We achieve it by providing performance information to identify GPU program performance bottlenecks and decide which optimization methods should be adopted, so as to facilitate the best match between algorithm features and underlying hardware characteristics. To demonstrate the usage of tuning chain, we optimize three representative GPU kernels with different compute intensity: Matrix Transpose, Laplace Transform and Integral on both NVIDIA and AMD GPUs. Experimental results demonstrate that under the guidance of our tuning chain, performance of those kernels achieves 7.8~42.4 times speedup compared to their nai¨ve implementations on both NVIDIA and AMD GPU platforms. © 2012 Springer-Verlag.; It is challenging to optimize GPU kernels because this progress requires deep technical knowledge of the underlying hardware. Modern GPU architectures are becoming more and more diversified, which further exacerbates the already difficult problem of performance optimization. This paper presents an insightful performance tuning chain for GPUs. The goal is to help non-expert programmers with limited knowledge of GPU architectures implement high performance GPU kernels directly. We achieve it by providing performance information to identify GPU program performance bottlenecks and decide which optimization methods should be adopted, so as to facilitate the best match between algorithm features and underlying hardware characteristics. To demonstrate the usage of tuning chain, we optimize three representative GPU kernels with different compute intensity: Matrix Transpose, Laplace Transform and Integral on both NVIDIA and AMD GPUs. Experimental results demonstrate that under the guidance of our tuning chain, performance of those kernels achieves 7.8~42.4 times speedup compared to their nai¨ve implementations on both NVIDIA and AMD GPU platforms. © 2012 Springer-Verlag.
KeywordHardware Laplace Transforms Optimization Program Processors
Language英语
Content Type会议论文
URIhttp://ir.iscas.ac.cn/handle/311060/15798
Collection中国科学院软件研究所
Recommended Citation
GB/T 7714
Jia Haipeng,Zhang Yunquan,Long Guoping,et al. an insightful program performance tuning chain for gpu computing[C],2012:502-516.
Files in This Item:
There are no files associated with this item.
Related Services
Recommend this item
Bookmark
Usage statistics
Export to Endnote
Google Scholar
Similar articles in Google Scholar
[Jia Haipeng]'s Articles
[Zhang Yunquan]'s Articles
[Long Guoping]'s Articles
Baidu academic
Similar articles in Baidu academic
[Jia Haipeng]'s Articles
[Zhang Yunquan]'s Articles
[Long Guoping]'s Articles
Bing Scholar
Similar articles in Bing Scholar
[Jia Haipeng]'s Articles
[Zhang Yunquan]'s Articles
[Long Guoping]'s Articles
Terms of Use
No data!
Social Bookmark/Share
All comments (0)
No comment.
 

Items in the repository are protected by copyright, with all rights reserved, unless otherwise indicated.