Implementation of Variable Preconditioned GCR with mixed precision on GPU using CUDA

Author

Ikuno, Soichiro ; Fujita, Norihisa ; Yamamoto, Susumu ; Nakata, Susumu

Author_Institution

Tokyo Univ. of Technol., Hachioji, Japan

fYear

2010

fDate

9-12 May 2010

Firstpage

1

Lastpage

1

Abstract

The Variable Preconditioned GVR (VPGCR) with mixed precision on Graphics Processing Unit (GPU) using Compute Unified Device Architecture (CUDA) is numerically investigated. The convergence theorem of VPGCR is guaranteed that the residual equation for the preconditioned procedure can be solved in the range of single precision operation. The results of computations show that VPGCR with mixed precision operation on GPU demonstrated significant achievement than that of CPU. Especially, VPGCR on GPU with mixed precision operation is 22.53 times faster than that of Central Processing Unit (CPU).

Keywords

computer graphics; convergence; coprocessors; parallel architectures; central processing unit; compute unified device architecture; convergence theorem; graphics processing unit; mixed precision; variable preconditioned GCR; Central Processing Unit; Computer architecture; Computer graphics; Concurrent computing; Equations; Hardware; Jacobian matrices; Linear systems; Scientific computing; Software development management;

fLanguage

English

Publisher

ieee

Conference_Titel

Electromagnetic Field Computation (CEFC), 2010 14th Biennial IEEE Conference on

Conference_Location

Chicago, IL

Print_ISBN

978-1-4244-7059-4

Type

conf

DOI

10.1109/CEFC.2010.5481534

Filename

5481534