The Optimization of FFT Algorithm Based with Parallel Computing on GPU

FFTW and CUFFT are used as typical FFT computing libraries based on CPU and GPU respectively. This paper tests and analyzes the performance and total consumption time of machine floating-point operation accelerated by CPU and GPU algorithm under the same data volume. The results show that CUFFT based on GPU has a better comprehensive performance than FFTW.

[1]  Jean-Luc Starck,et al.  Astronomical image and data analysis , 2002 .

[2]  Kenneth Moreland,et al.  The FFT on a GPU , 2003, HWWS '03.

[3]  Mikhail Kapralov,et al.  Sparse fourier transform in any constant dimension with nearly-optimal sample complexity in sublinear time , 2016, STOC.

[4]  Xiaoming Li,et al.  A hybrid GPU/CPU FFT library for large FFT problems , 2013, 2013 IEEE 32nd International Performance Computing and Communications Conference (IPCCC).

[5]  Zhiyi Yang,et al.  Parallel Image Processing Based on CUDA , 2008, 2008 International Conference on Computer Science and Software Engineering.

[6]  Piotr Indyk,et al.  (Nearly) Sample-Optimal Sparse Fourier Transform , 2014, SODA.

[7]  Lister Staveley-Smith,et al.  GPU accelerated radio astronomy signal convolution , 2008 .

[8]  Steven G. Johnson,et al.  The Fastest Fourier Transform in the West , 1997 .

[9]  E. Brigham,et al.  The fast Fourier transform and its applications , 1988 .