mpfft: an auto-tuning fft library for opencl gpus

ISCAS OpenIR

	mpfft: an auto-tuning fft library for opencl gpus
	Li Yan; Zhang Yun-Quan; Liu Yi-Qun; Long Guo-Ping; Jia Hai-Peng
	2013
会议录名称	Journal of Computer Science and Technology
页码	90-105
收录类别	EI
ISSN	1000-9000
部门归属	(1) Institute of Software Chinese Academy of Sciences Beijing 100190 China; (2) Graduate University of Chinese Academy of Sciences Beijing 100049 China; (3) School of Information Science and Engineering Ocean University of China Qingdao 266000 China
摘要	Fourier methods have revolutionized many fields of science and engineering, such as astronomy, medical imaging, seismology and spectroscopy, and the fast Fourier transform (FFT) is a computationally efficient method of generating a Fourier transform. 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 software. However, the complexity of GPU programming poses a significant challenge to developers. In this paper, we propose an automatic performance tuning framework for FFT on various OpenCL GPUs, and implement a high performance library named MPFFT based on this framework. For power-of-two length FFTs, our library substantially outperforms the clAmdFft library on AMD GPUs and achieves comparable performance as the CUFFT library on NVIDIA GPUs. Furthermore, our library also supports non-power-of-two size. For 3D non-power-of-two FFTs, our library delivers 1.5x to 28x faster than FFTW with 4 threads and 20.01x average speedup over CUFFT 4.0 on Tesla C2050. © 2013 Springer Science+Business Media New York & Science Press, China.; Fourier methods have revolutionized many fields of science and engineering, such as astronomy, medical imaging, seismology and spectroscopy, and the fast Fourier transform (FFT) is a computationally efficient method of generating a Fourier transform. 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 software. However, the complexity of GPU programming poses a significant challenge to developers. In this paper, we propose an automatic performance tuning framework for FFT on various OpenCL GPUs, and implement a high performance library named MPFFT based on this framework. For power-of-two length FFTs, our library substantially outperforms the clAmdFft library on AMD GPUs and achieves comparable performance as the CUFFT library on NVIDIA GPUs. Furthermore, our library also supports non-power-of-two size. For 3D non-power-of-two FFTs, our library delivers 1.5x to 28x faster than FFTW with 4 threads and 20.01x average speedup over CUFFT 4.0 on Tesla C2050. © 2013 Springer Science+Business Media New York & Science Press, China.
关键词	Medical Imaging Program Processors
语种	英语
内容类型	会议论文
URI标识	http://ir.iscas.ac.cn/handle/311060/15976
专题	中国科学院软件研究所
推荐引用方式 GB/T 7714	Li Yan,Zhang Yun-Quan,Liu Yi-Qun,et al. mpfft: an auto-tuning fft library for opencl gpus[C],2013:90-105.