an insightful program performance tuning chain for gpu computing
文献类型：会议论文
| 作者 | Jia Haipeng ; Zhang Yunquan ; Long Guoping ; Yan Shengen |
| 出版日期 | 2012 |
| 会议名称 | 12th International Conference on Algorithms and Architectures for Parallel Processing, ICA3PP 2012 |
| 会议日期 | September 4, 2012 - September 7, 2012 |
| 会议地点 | Fukuoka, Japan |
| 关键词 | Hardware Laplace transforms Optimization Program processors |
| 页码 | 502-516 |
| 中文摘要 | 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. |
| 英文摘要 | 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. |
| 收录类别 | EI |
| 会议录 | Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
 |
| 语种 | 英语 |
| ISSN号 | 0302-9743 |
| ISBN号 | 9783642330773 |
| 源URL | [http://ir.iscas.ac.cn/handle/311060/15798]  |
| 专题 | 软件研究所_软件所图书馆_会议论文 |
| 推荐引用方式 GB/T 7714 | Jia Haipeng,Zhang Yunquan,Long Guoping,et al. an insightful program performance tuning chain for gpu computing[C]. 见:12th International Conference on Algorithms and Architectures for Parallel Processing, ICA3PP 2012. Fukuoka, Japan. September 4, 2012 - September 7, 2012. |
入库方式： OAI收割
来源：软件研究所
浏览0
下载0
收藏0
其他版本
除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。