Novel Research on a Finite-Difference Time-Domain Acceleration Algorithm Based on Distributed Cluster Graphic Process Units

In computational electromagnetics, the finite-difference time-domain (FDTD) method is recognized for its volumetric discretization approach. However, it can be computationally demanding when addressing large-scale electromagnetic problems. This paper introduces a novel approach by incorporating Grap...

Full description

Saved in:
Bibliographic Details
Main Authors: Xinbo He, Shenggang Mu, Xudong Han, Bing Wei
Format: Article
Language:English
Published: MDPI AG 2025-04-01
Series:Applied Sciences
Subjects:
Online Access:https://www.mdpi.com/2076-3417/15/9/4834
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850137744870735872
author Xinbo He
Shenggang Mu
Xudong Han
Bing Wei
author_facet Xinbo He
Shenggang Mu
Xudong Han
Bing Wei
author_sort Xinbo He
collection DOAJ
description In computational electromagnetics, the finite-difference time-domain (FDTD) method is recognized for its volumetric discretization approach. However, it can be computationally demanding when addressing large-scale electromagnetic problems. This paper introduces a novel approach by incorporating Graphic Process Units (GPUs) into an FDTD algorithm. It leverages the Compute Unified Device Architecture (CUDA) along with OpenMPI and the NVIDIA Collective Communications Library (NCCL) to establish a parallel scheme for the FDTD algorithm in distributed cluster GPUs. This approach enhances the computational efficiency of the FDTD algorithm by circumventing data relaying by the CPU and the limitations of the PCIe bus. The improved efficiency renders the FDTD algorithm a more practical and efficient solution for real-world electromagnetic problems.
format Article
id doaj-art-ea3e9eb471e545efba452fa99c4441de
institution OA Journals
issn 2076-3417
language English
publishDate 2025-04-01
publisher MDPI AG
record_format Article
series Applied Sciences
spelling doaj-art-ea3e9eb471e545efba452fa99c4441de2025-08-20T02:30:45ZengMDPI AGApplied Sciences2076-34172025-04-01159483410.3390/app15094834Novel Research on a Finite-Difference Time-Domain Acceleration Algorithm Based on Distributed Cluster Graphic Process UnitsXinbo He0Shenggang Mu1Xudong Han2Bing Wei3School of Physics, Xidian University, Xi’an 710071, ChinaSchool of Physics, Xidian University, Xi’an 710071, ChinaSchool of Physics, Xidian University, Xi’an 710071, ChinaSchool of Physics, Xidian University, Xi’an 710071, ChinaIn computational electromagnetics, the finite-difference time-domain (FDTD) method is recognized for its volumetric discretization approach. However, it can be computationally demanding when addressing large-scale electromagnetic problems. This paper introduces a novel approach by incorporating Graphic Process Units (GPUs) into an FDTD algorithm. It leverages the Compute Unified Device Architecture (CUDA) along with OpenMPI and the NVIDIA Collective Communications Library (NCCL) to establish a parallel scheme for the FDTD algorithm in distributed cluster GPUs. This approach enhances the computational efficiency of the FDTD algorithm by circumventing data relaying by the CPU and the limitations of the PCIe bus. The improved efficiency renders the FDTD algorithm a more practical and efficient solution for real-world electromagnetic problems.https://www.mdpi.com/2076-3417/15/9/4834finite-difference time domaingraphic process unitsdistributed clusterNVIDIA Collective Communication Library
spellingShingle Xinbo He
Shenggang Mu
Xudong Han
Bing Wei
Novel Research on a Finite-Difference Time-Domain Acceleration Algorithm Based on Distributed Cluster Graphic Process Units
Applied Sciences
finite-difference time domain
graphic process units
distributed cluster
NVIDIA Collective Communication Library
title Novel Research on a Finite-Difference Time-Domain Acceleration Algorithm Based on Distributed Cluster Graphic Process Units
title_full Novel Research on a Finite-Difference Time-Domain Acceleration Algorithm Based on Distributed Cluster Graphic Process Units
title_fullStr Novel Research on a Finite-Difference Time-Domain Acceleration Algorithm Based on Distributed Cluster Graphic Process Units
title_full_unstemmed Novel Research on a Finite-Difference Time-Domain Acceleration Algorithm Based on Distributed Cluster Graphic Process Units
title_short Novel Research on a Finite-Difference Time-Domain Acceleration Algorithm Based on Distributed Cluster Graphic Process Units
title_sort novel research on a finite difference time domain acceleration algorithm based on distributed cluster graphic process units
topic finite-difference time domain
graphic process units
distributed cluster
NVIDIA Collective Communication Library
url https://www.mdpi.com/2076-3417/15/9/4834
work_keys_str_mv AT xinbohe novelresearchonafinitedifferencetimedomainaccelerationalgorithmbasedondistributedclustergraphicprocessunits
AT shenggangmu novelresearchonafinitedifferencetimedomainaccelerationalgorithmbasedondistributedclustergraphicprocessunits
AT xudonghan novelresearchonafinitedifferencetimedomainaccelerationalgorithmbasedondistributedclustergraphicprocessunits
AT bingwei novelresearchonafinitedifferencetimedomainaccelerationalgorithmbasedondistributedclustergraphicprocessunits