automatic fft performance tuning on opencl gpus

ISCAS OpenIR

	automatic fft performance tuning on opencl gpus
	Li Yan; Zhang Yunquan; Jia Haipeng; Long Guoping; Wang Ke
	2011
会议名称	2011 17th IEEE International Conference on Parallel and Distributed Systems, ICPADS 2011
会议录名称	Proceedings of the International Conference on Parallel and Distributed Systems - ICPADS
页码	228-235
会议日期	December 7, 2011 - December 9, 2011
会议地点	Tainan, Taiwan
收录类别	EI
ISSN	1521-9097
ISBN	9780769545769
部门归属	(1) Laboratory of Parallel Software and Computational Science Institute of Software Chinese Academy of Sciences Beijing China; (2) State Key Lab. of Computer Science Institute of Software Chinese Academy of Sciences Beijing China; (3) Chinese Academy of Sciences Graduate University Beijing China; (4) School of Information Science and Engineering Ocean University Qingdao China
摘要	School of Information Science and Engineering, Ocean University of China, Qingdao, China Many fields of science and engineering, such as astronomy, medical imaging, seismology and spectroscopy, have been revolutionized by Fourier methods. The fast Fourier transform (FFT) is an efficient algorithm to compute the discrete Fourier transform (DFT) and its inverse. The emerging class of high performance computing architectures, such as GPU, seeks to achieve much higher performance and efficiency by exposing a hierarchy of distinct memories to programmers. However, the complexity of GPU programming poses a significant challenge for programmers. In this paper, based on the Kronecker product form multi-dimensional FFTs, we propose an automatic performance tuning framework for various OpenCL GPUs. Several key techniques of GPU programming on AMD and NVIDIA GPUs are also identified. Our OpenCL FFT library achieves up to 1.5 to 4 times, 1.5 to 40 times and 1.4 times the performance of clAmdFft 1.0 for 1D, 2D and 3D FFT respectively on an AMD GPU, and the overall performance is within 90% of CUFFT 4.0 on two NVIDIA GPUs. © 2011 IEEE.; School of Information Science and Engineering, Ocean University of China, Qingdao, China Many fields of science and engineering, such as astronomy, medical imaging, seismology and spectroscopy, have been revolutionized by Fourier methods. The fast Fourier transform (FFT) is an efficient algorithm to compute the discrete Fourier transform (DFT) and its inverse. The emerging class of high performance computing architectures, such as GPU, seeks to achieve much higher performance and efficiency by exposing a hierarchy of distinct memories to programmers. However, the complexity of GPU programming poses a significant challenge for programmers. In this paper, based on the Kronecker product form multi-dimensional FFTs, we propose an automatic performance tuning framework for various OpenCL GPUs. Several key techniques of GPU programming on AMD and NVIDIA GPUs are also identified. Our OpenCL FFT library achieves up to 1.5 to 4 times, 1.5 to 40 times and 1.4 times the performance of clAmdFft 1.0 for 1D, 2D and 3D FFT respectively on an AMD GPU, and the overall performance is within 90% of CUFFT 4.0 on two NVIDIA GPUs. © 2011 IEEE.
关键词	Algorithms Computer Systems Discrete Fourier Transforms Fast Fourier Transforms Medical Imaging
主办者	National Cheng Kung University; National Science Council; Ministry of Education; Academia Sinica; National Center for High Performance Computing
语种	英语
内容类型	会议论文
URI标识	http://ir.iscas.ac.cn/handle/311060/16294
专题	中国科学院软件研究所
推荐引用方式 GB/T 7714	Li Yan,Zhang Yunquan,Jia Haipeng,et al. automatic fft performance tuning on opencl gpus[C],2011:228-235.