An implementation of parallel 3-d FFT using short vector SIMD instructions on clusters of PCs

Authors:
Daisuke Takahashi;Taisuke Boku;Mitsuhisa Sato
Affiliations:
Graduate School of Systems and Information Engineering, University of Tsukuba, Tsukuba, Ibaraki, Japan;Graduate School of Systems and Information Engineering, University of Tsukuba, Tsukuba, Ibaraki, Japan;Graduate School of Systems and Information Engineering, University of Tsukuba, Tsukuba, Ibaraki, Japan
Venue:
PARA'04 Proceedings of the 7th international conference on Applied Parallel Computing: state of the Art in Scientific Computing
Year:
2004

Citing 9
Cited 0

Two and three dimensional FFTS on highly parallel computers

Parallel Computing
FFTs in external or hierarchical memory

The Journal of Supercomputing
Computational frameworks for the fast Fourier transform

Computational frameworks for the fast Fourier transform
Real and complex fast Fourier transforms on the Fujitsu VPP 500

Parallel Computing
Implementation of parallel FFT algorithms on distributed memory machines with a minimum overhead of communication

Parallel Computing
High Performance Communication using a Commodity Network for Cluster Systems

HPDC '00 Proceedings of the 9th IEEE International Symposium on High Performance Distributed Computing
Radix-4 FFT implementation using SIMD multimedia instructions

ICASSP '99 Proceedings of the Acoustics, Speech, and Signal Processing, 1999. on 1999 IEEE International Conference - Volume 04
Architecture independent short vector FFTs

ICASSP '01 Proceedings of the Acoustics, Speech, and Signal Processing, 200. on IEEE International Conference - Volume 02
FFT algorithms for vector computers

Parallel Computing

Quantified Score

Hi-index	0.00

Visualization

Abstract

In this paper, we propose an implementation of a parallel three-dimensional fast Fourier transform (FFT) using short vector SIMD instructions on clusters of PCs. We vectorized FFT kernels using Intel's Streaming SIMD Extensions 2 (SSE2) instructions. We show that a combination of the vectorization and block three-dimensional FFT algorithm improves performance effectively. Performance results of three-dimensional FFTs on a dual Xeon 2.8 GHz PC SMP cluster are reported. We successfully achieved performance of over 5 GFLOPS on a 16-node dual Xeon 2.8 GHz PC SMP cluster.