DocumentCode
2682908
Title
The Graphics Processor as a Mathematical Coprocessor in MATLAB
Author
Brodtkorb, André Rigland
Author_Institution
Dept. of Appl. Math., SINTEF ICT, Oslo
fYear
2008
fDate
4-7 March 2008
Firstpage
822
Lastpage
827
Abstract
We present an interface to the graphics processing unit (GPU) from MATLAB, and four algorithms from numerical linear algebra available through this interface; matrix-matrix multiplication, Gauss-Jordan elimination, PLU factorization, and tridiagonal Gaussian elimination. In addition to being a high level abstraction to the GPU, the interface offers background processing, enabling computations to be executed on the CPU simultaneously. The algorithms are shown to be up-to 31 times faster than highly optimized CPU code. The algorithms have only been tested on single precision hardware, but will easily run on new double precision hardware.
Keywords
coprocessors; linear algebra; mathematics computing; Gauss-Jordan elimination; MATLAB; PLU factorization; graphics processing unit; graphics processor; high- level abstraction; mathematical coprocessor; matrix-matrix multiplication; numerical linear algebra; tridiagonal Gaussian elimination; Central Processing Unit; Commercialization; Computer graphics; Coprocessors; Gaussian processes; Hardware; Linear algebra; MATLAB; Mathematical model; Rendering (computer graphics); Background Computation; GPU; MATLAB; Numerical Linear Algebra;
fLanguage
English
Publisher
ieee
Conference_Titel
Complex, Intelligent and Software Intensive Systems, 2008. CISIS 2008. International Conference on
Conference_Location
Barcelona
Print_ISBN
978-0-7695-3109-0
Type
conf
DOI
10.1109/CISIS.2008.68
Filename
4606774
Link To Document