Implementation of Parallel 1-D FFT on GPU Clusters

Implementation of Parallel 1-D FFT on GPU Clusters
复制标题

DOI:
10.1109/cse.2013.36
复制
发表时间:
2013-12
期刊:
2013 IEEE 16th International Conference on Computational Science and Engineering
影响因子:
--
通讯作者:
D. Takahashi
D. Takahashi
中科院分区:
其他
文献类型:
--
作者:
D. Takahashi

文献摘要

被引文献

相似文献

在本文中,我们提出了一个并行的一维快速傅立叶变换(FFT)的GPU集群上的实现。该实现基于六步FFT算法。由于并行一维FFT需要三次全对全通信,因此GPU集群上的并行FFT的一个目标是最小化PCI Express传输时间和MPI通信时间。我们证明了MVAPICH 2-GPU的高级功能可以轻松地重叠PCI Express传输和MPI通信。一维FFT的GPU集群上的性能结果的报告。对于234点FFT,我们在HA-PACS的128个节点上成功实现了超过763 GFlops的性能(268个节点,2.99 TFlops/节点,802 TFlops峰值性能)。
In this paper, we propose an implementation of a parallel one-dimensional fast Fourier transform (FFT) on GPU clusters. This implementation is based on the six-step FFT algorithm. Because the parallel one-dimensional FFT requires three all-to-all communications, one goal for parallel FFTs on GPU clusters is to minimize the PCI Express transfer time and the MPI communication time. We demonstrate that the advanced features of MVAPICH2-GPU make it easy to overlap PCI Express transfers and MPI communication. Performance results of one-dimensional FFTs on a GPU cluster are reported. We successfully achieved a performance of over 763 GFlops on 128 nodes of the HA-PACS (268 nodes, 2.99 TFlops/node, 802 TFlops peak performance) for 234-point FFT.