Minimizing Communication in the Multidimensional FFT
SIAM Journal on Scientific Computing · 31.12.2023 · 45 (6) · C330-C347
DOI: 10.1137/22m1487242 ↗Kurz gefasst
Fachartikel in der Zeitschrift SIAM Journal on Scientific Computing, online erschienen am 08.12.2023. Verfasst von Thomas Koopman und 1 weiteren Autor:innen. Die Arbeit wurde bisher 6-mal zitiert (Crossref) und nennt 26 Literaturangaben. Gefördert von: Dutch Research Council.
Abstract (Originalsprache)
We present a parallel algorithm for the fast Fourier transform (FFT) in higher dimensions. This algorithm generalizes the cyclic-to-cyclic one-dimensional parallel algorithm to a cyclic-to-cyclic multidimensional parallel algorithm while retaining the property of needing only a single all-to-all communication step. This is under the constraint that we use at most [Formula: see text] processors for an FFT on an array with a total of [Formula: see text] elements, irrespective of the dimension [Formula: see text] or the shape of the array. The only assumption we make is that [Formula: see text] is sufficiently composite. Our algorithm starts and ends in the same data distribution. We present our multidimensional implementation FFTU which utilizes the sequential FFTW program for its local FFTs, and which can handle any dimension [Formula: see text]. We obtain experimental results for [Formula: see text] using MPI on up to 4096 cores of the supercomputer Snellius, comparing FFTU with the parallel FFTW program and with PFFT and heFFTe. These results show that FFTU is competitive with the state of the art and that it allows one to use a larger number of processors, while keeping communication limited to a single all-to-all operation. For arrays of size [Formula: see text] and [Formula: see text], FFTU achieves a speedup of a factor 149 and 176, respectively, on 4096 processors.
Förderung
- Dutch Research Council
Zitieren
APA
Koopman, T., & Bisseling, R. H. (2023). Minimizing Communication in the Multidimensional FFT. SIAM Journal on Scientific Computing, 45(6), C330-C347. https://doi.org/10.1137/22m1487242
Vancouver
Koopman T, Bisseling RH. Minimizing Communication in the Multidimensional FFT. SIAM Journal on Scientific Computing. 2023;45(6):C330-C347. doi:10.1137/22m1487242
BibTeX
@article{Koopman2023,
author = {Koopman, Thomas and Bisseling, Rob H.},
title = {Minimizing Communication in the Multidimensional FFT},
journal = {SIAM Journal on Scientific Computing},
year = {2023},
volume = {45},
number = {6},
pages = {C330--C347},
doi = {10.1137/22m1487242}
}
EUVIMED