Implementation of Parallel 3-D Real FFT with 2-D Decomposition on Intel Xeon Phi Clusters

Implementation of Parallel 3-D Real FFT with 2-D Decomposition on Intel Xeon Phi Clusters
复制标题

在英特尔至强融核集群上实现并行 3-D Real FFT 和 2-D 分解

DOI:
10.1007/978-3-030-43229-4_14
复制
发表时间:
2020
期刊:
Proc. 13th International Conference on Parallel Processing and Applied Mathematics (PPAM 2019), Part I, Lecture Notes in Computer Science
影响因子:
--
通讯作者:
Daisuke Takahashi
Daisuke Takahashi
中科院分区:
--
文献类型:
--
作者:
齋藤光貴;渡辺知恵美;Daisuke Takahashi

文献摘要

相似文献

在本文中,我们提出了一个并行的3-D真实的快速傅立叶变换(FFT)与2-D分解的英特尔至强融核集群的实现。基于离散傅里叶变换(DFT)的共轭对称性和行列FFT算法,提出了并行三维真实的FFT的实现方法。我们使用英特尔高级矢量扩展512(英特尔AVX-512)指令对FFT内核进行矢量化。报告了英特尔至强融核群集上并行3-D真实的FFT的性能结果。在富士通PRIMERGY CX 1640 M1集群的2048个节点上,我们成功地实现了一点FFT的10 TFlops以上的性能水平。
In this paper, we propose an implementation of a parallel 3-D real fast Fourier transform (FFT) with 2-D decomposition on Intel Xeon Phi clusters. The proposed implementation of the parallel 3-D real FFT is based on the conjugate symmetry property of the discrete Fourier transform (DFT) and the row-column FFT algorithm. We vectorized FFT kernels using the Intel Advanced Vector Extensions 512 (Intel AVX-512) instructions. Performance results of parallel 3-D real FFTs on Intel Xeon Phi clusters are reported. We successfully achieved a level of performance over 10 TFlops on 2048 nodes of Fujitsu PRIMERGY CX1640 M1 cluster for an-point FFT.