An Iterative Parallel Solver in GPU Applied to Frequency Domain Linear Water Wave Problems by the Boundary Element Method

Frontiers in Built Environment

nez-Castro, Alejandro Enrique¹  Molina-Moya, Jorge¹  Martí²  Ortiz, Pablo² 

[1] Department of Structural Mechanics and Hydraulic Engineering, University of Granada, Edificio Politécnico, Spain

关键词: GMRES (generalized minimal residual) algorithm;    CUDA (compute unified device architecture);    GPU (CUDA);    Floating bodies;    Boundary Element Method - BEM;   

DOI  :  10.3389/fbuil.2018.00069

学科分类：建筑学

来源: Frontiers

【 摘 要 】

In this paper the implementation of an iterative solver based on the Generalized Minimum Residual Method (GMRES) with complex-valued coefficients is explored, with application to the Boundary Element Method (BEM). The solver is designed to be implemented in a GPU (Graphic Processing Unit) device, exploiting its massively parallel capabilities. The framework is in the context of linear water wave problems in the frequency domain. The main objective of the proposed solver is the direct acceleration of existing standard BEM codes, by transfering to the GPU the solver task. The CUDA programming language is used, exploiting the particular architecture of the GPU device for complex-valued systems. To explore the performances of the solver, two linear water wave problems have been tested: the frequency-dependent added mass and damping matrices of a 3D floating body, and the Helmholtz equation in a 2D domain. The parallelized GMRES solver is implemented in a NVidia GeForce GTX 970 graphic card, and shows drastic reductions in computing times when compared with its CPU implementation.

【 授权许可】

CC BY