Parallel SDDM with communication-avoiding BICGSTAB for large-scale 3D scattering problems

Ming Ye Fan, Rui Qing Liu, Ming Lin Yang*, Xin Qing Sheng

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

In the parallel finite element domain decomposition methods (FEM-DDM), one of the key step is to solve the reduced linear system defined on the subdomain interfaces iteratively using a Krylov subspace method such as the biconjugate gradient stabilized (BICGSTAB) method. In each iteration of BICGSTAB, multiple global MPI communication collectives are required, which are becoming increasingly expensive when many processors are used on modern computers, especially on a large-scale system. To address this limitation, a communication-avoiding variant of BICGSTAB (CA-BICGSTAB) is realized and implemented to minimize global communication therefore improve massively parallel efficiency of the FEM-DDM in this paper. A series of numerical examples are presented to demonstrate the superior performance of CA-BICGSTAB over conventional BICGSTAB.

Original languageEnglish
Title of host publication2019 International Applied Computational Electromagnetics Society Symposium-China, ACES 2019
PublisherInstitute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)9780996007894
DOIs
Publication statusPublished - Aug 2019
Event2019 International Applied Computational Electromagnetics Society Symposium-China, ACES 2019 - Nanjing, China
Duration: 8 Aug 201911 Aug 2019

Publication series

Name2019 International Applied Computational Electromagnetics Society Symposium-China, ACES 2019

Conference

Conference2019 International Applied Computational Electromagnetics Society Symposium-China, ACES 2019
Country/TerritoryChina
CityNanjing
Period8/08/1911/08/19

Keywords

  • Communication avoiding BICGSTAB
  • Finite element domain decomposition method
  • Parallel
  • Scattering

Fingerprint

Dive into the research topics of 'Parallel SDDM with communication-avoiding BICGSTAB for large-scale 3D scattering problems'. Together they form a unique fingerprint.

Cite this