A New Parallel Frequency-Domain Finite-Difference Algorithm Using Multi-GPU

被引：0

作者：

Wang, Yijing ^{[1
]}

He, Xinbo ^{[1
]}

Wei, Bin ^{[1
]}

机构：

[1] Xidian Univ, Sch Phys, Xian 710071, Peoples R China

来源：

IEEE MICROWAVE AND WIRELESS TECHNOLOGY LETTERS | 2024年 / 34卷 / 08期

基金：

中国国家自然科学基金;

关键词：

Graphics processing units; Sparse matrices; Computational efficiency; Finite difference methods; Matrix decomposition; Instruction sets; Mathematical models; Electromagnetic scattering; frequency-domain finite-difference (FDFD); graphic processing unit (GPU); multi-GPU; parallel algorithm; CONJUGATE GRADIENTS; ACCELERATION; FDTD;

D O I：

10.1109/LMWT.2024.3414598

中图分类号：

TM [电工技术]; TN [电子技术、通信技术];

学科分类号：

0808 ; 0809 ;

摘要：

This letter presents a parallel frequency-domain finite-difference (FDFD) algorithm based on multi-graphic processing unit (GPU) applied to electromagnetic scattering computations to enhance the computational efficiency of the algorithm. The proposed algorithm parallelizes the solution of large-scale sparse matrices, distributing threads to the matrix-vector and vector-vector multiplication operations within decomposed sub-matrices to reduce the computational time. Moreover, we configure the OpenMP to optimize communication transfer between multiple GPUs, thereby improving computational efficiency. The simulation results show that compared with the conventional FDFD method, the proposed algorithm can enhance computational efficiency while ensuring accuracy.

引用

页码：971 / 974

页数：4

共 50 条

[31] Finite-Difference Complex-Frequency-Domain Method for Optical and Plasmonic Analyses
Wu, Di
Ohnishi, R.
Uemura, R.
Yamaguchi, T.
Ohnuki, S.
IEEE PHOTONICS TECHNOLOGY LETTERS, 2018, 30 (11) : 1024 - 1027
[32] A new domain decomposition finite-difference time domain for solving large electromagnetic problems
Abd-El-Raouf, Hany E.
Mittra, Raj
Ma, Ji-Fu
MICROWAVE AND OPTICAL TECHNOLOGY LETTERS, 2006, 48 (12) : 2399 - 2405
[33] A Novel Scheme for High Performance Finite-Difference Time-Domain (FDTD) Computations Based on GPU
Chu, Tianshu
Dai, Jian
Qian, Depei
Fang, Weiwei
Liu, Yi
ALGORITHMS AND ARCHITECTURES FOR PARALLEL PROCESSING, PT 1, PROCEEDINGS, 2010, 6081 : 441 - +
[34] GPU-accelerated finite-difference time-domain method for dielectric media based on CUDA
Wang, Ximin
Liu, Song
Li, Xuan
Zhong, Shuangying
INTERNATIONAL JOURNAL OF RF AND MICROWAVE COMPUTER-AIDED ENGINEERING, 2016, 26 (06) : 512 - 518
[35] ELECTROMAGNETIC MODELING USING THE FINITE-DIFFERENCE TIME-DOMAIN METHOD
DUCEAU, E
RECHERCHE AEROSPATIALE, 1994, (05): : 301 - 317
[36] Fast Parallel Computation of Computer-Generated Hologram using a multi-GPU environmental PC
Takada, Naoki
Shimobaba, Tomoyoshi
Sugiyama, Atsushi
Okada, Naohisa
Nakayama, Hirotaka
Shiraki, Atsushi
Masuda, Nobuyuki
Ito, Tomoyoshi
IDW/AD '12: PROCEEDINGS OF THE INTERNATIONAL DISPLAY WORKSHOPS, PT 2, 2012, 19 : 1281 - 1282
[37] Using MATLAB's Parallel Processing Toolbox for Multi-CPU and Multi-GPU Accelerated FDTD Simulations
Weiss, Alec J.
Elsherbeni, Atef Z.
Demir, Veysel
Hadi, Mohammed F.
APPLIED COMPUTATIONAL ELECTROMAGNETICS SOCIETY JOURNAL, 2019, 34 (05): : 724 - 730
[38] Multi-GPU Based Parallel Design of the Ant Colony Optimization Algorithm for Endmember Extraction from Hyperspectral Images
Gao, Jianwei
Sun, Yi
Zhang, Bing
Chen, Zhengchao
Gao, Lianru
Zhang, Wenjuan
SENSORS, 2019, 19 (03):
[39] Hardware Implementation of a Three-Dimensional Finite-Difference Time-Domain Algorithm
Durbano, James P.
Ortiz, Fernando E.
Humphrey, John R.
Mirotznik, Mark S.
Prather, Dennis W.
IEEE ANTENNAS AND WIRELESS PROPAGATION LETTERS, 2003, 2 (02): : 54 - 57
[40] Huygens Subgridding for 3-D Frequency-Dependent Finite-Difference Time-Domain Method
Abalenkovs, Maksims
Costen, Fumie
Berenger, Jean-Pierre
Himeno, Ryutaro
Yokota, Hideo
Fujii, Masafumi
IEEE TRANSACTIONS ON ANTENNAS AND PROPAGATION, 2012, 60 (09) : 4336 - 4344

← 1 2 3 4 5 →