Title :
Parallel Implementation and Performance Comparison of BiCGStab for Massive Sparse Linear System of Equations on GPU Libraries
Author :
Liankai Yao;Xiaohui Ji;Shuai Liu;Jiayue Yang
Author_Institution :
Sch. of Inf. Eng., China Univ. of Geosci. (Beijing), Beijing, China
Abstract :
Biconjugate Gradient Stabilized (BiCGStab) method is a popular iterative method for solving massive linear system of equations with sparse asymmetric coefficient matrix. Implementing methods of BiCGStab on CUDA-enabled GPU libraries are given in the paper. Performance comparisons are made. Different matrix storage formats were used in the implementations and the experiments. Based on the study, suggestions of solving different kinds of linear system of equations are given.
Keywords :
"Sparse matrices","Libraries","Graphics processing units","Mathematical model","Linear systems","Kernel","Parallel processing"
Conference_Titel :
Ubiquitous Intelligence and Computing and 2015 IEEE 12th Intl Conf on Autonomic and Trusted Computing and 2015 IEEE 15th Intl Conf on Scalable Computing and Communications and Its Associated Workshops (UIC-ATC-ScalCom), 2015 IEEE 12th Intl Conf on
DOI :
10.1109/UIC-ATC-ScalCom-CBDCom-IoP.2015.119
