FINITE ELEMENT MATRIX GENERATION ON A GPU

被引：59

作者：

Dziekonski, A. ^{[1
]}

Sypek, P. ^{[1
]}

Lamecki, A. ^{[1
]}

Mrozowski, M. ^{[1
]}

机构：

[1] Gdansk Univ Technol, Wireless Commun Engn WiComm Ctr Excellence, Dept Microwave & Antenna Engn, Fac Elect Telecommun & Informat,CUDA Res Ctr Comp, PL-80233 Gdansk, Poland

来源：

PROGRESS IN ELECTROMAGNETICS RESEARCH-PIER | 2012年 / 128卷

关键词：

TRANSMISSION CHARACTERISTICS; FEM; SCATTERING; EFFICIENT; DOMAIN; MODEL;

D O I：

10.2528/PIER12040301

中图分类号：

TM [电工技术]; TN [电子技术、通信技术];

学科分类号：

080906 [电磁信息功能材料与结构]; 082806 [农业信息与电气工程];

摘要：

This paper presents an efficient technique for fast generation of sparse systems of linear equations arising in computational electromagnetics in a finite element method using higher order elements. The proposed approach employs a graphics processing unit (GPU) for both numerical integration and matrix assembly. The performance results obtained on a test platform consisting of a Fermi GPU (1x Tesla C2075) and a CPU (2x twelve-core Opterons), indicate that the GPU implementation of the matrix generation allows one to achieve speedups by a factor of 81 and 19 over the optimized single- and multi-threaded CPU-only implementations, respectively.

引用

页码：249 / 265

页数：17

共 27 条

[1]

[Anonymous], 2011, CUDA EXAMPLE INTRO G

[2]

IMPROVING THREE-DIMENSIONAL ELECTRICAL CAPACITANCE TOMOGRAPHY IMAGING USING APPROXIMATION ERROR MODEL THEORY [J].