High-Performance Simulation of Electromagnetic Scattering by 3D Objects Using the GPU-accelerated Parallel MLFMA

Wei Jia He, Xin Duo Liu, Bi Yi Wu, Ming Lin Yang, Xin Qing Sheng

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

1 Citation (Scopus)

Abstract

This paper proposes a high-performance massively parallel multilevel fast multipole algorithm (MLFMA) on clusters with graphics processing units (GPUs) accelerators. ELLPACK format-based schemes are proposed to ensure a high computational throughput for GPU accelerated evaluation of interpolation and anterpolation operations in MLFMA. The program is further optimized using the mechanism of stream and the shared-memory/register hierarchical memory architecture of GPU. The implementation of all the proposed techniques make the aggregation/disaggregation calculations an order of magnitude faster than the conventional ones, and halved the iteration time for an aircraft model with 10 billion unknowns, which is a significant progress in GPU accelerated massively parallel MLFMA.

Original languageEnglish
Title of host publication2023 International Applied Computational Electromagnetics Society Symposium, ACES-China 2023
PublisherInstitute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)9781733509657
DOIs
Publication statusPublished - 2023
Event2023 International Applied Computational Electromagnetics Society Symposium, ACES-China 2023 - Hangzhou, China
Duration: 15 Aug 202318 Aug 2023

Publication series

Name2023 International Applied Computational Electromagnetics Society Symposium, ACES-China 2023

Conference

Conference2023 International Applied Computational Electromagnetics Society Symposium, ACES-China 2023
Country/TerritoryChina
CityHangzhou
Period15/08/2318/08/23

Keywords

  • GPU
  • Multilevel fast multipole algorithm
  • large-scale electromagnetic scattering
  • parallel

Fingerprint

Dive into the research topics of 'High-Performance Simulation of Electromagnetic Scattering by 3D Objects Using the GPU-accelerated Parallel MLFMA'. Together they form a unique fingerprint.

Cite this