Benchmarking of GPU NVIDIA CUDA, CUBLAS and MAGMA Libraries Based on Matrix Multiplication Problem

Authors

  • Edita E. Gichunts Institute for Informatics and Automation Problems of NAS RA

Keywords:

Parallel computing, Matrix algebra, Graphical processor

Abstract

Solving linear systems of equations is a fundamental problem in scientific computing. Many scientific computer applications need a high-performance matrix algebra. The major hardware developments always influenced the new developments in linear algebra libraries. Nowadays major chip manufacturers are developing next-generation microprocessor designs that integrate multicore CPU and GPU components [1]. The main aim is to benchmark CUBLAS and MAGMA libraries on matrix multiplication problem using the Tesla C1060 graphical processing unit.

References

L. Seiler, D. Carmean, E. Sprangle, T. Forsyth, M. Abrash, P. Dubey, S. Junkins, A. Lake, J. Sugerman, R. Cavin, R. Espasa, E. Grochowski, T. Juan and P. Hanrahan, “Larrabee: a many-core 86 architecture for visual computing”, ACM Trans. Graph., vol. 27, no. 3, pp. 1–15, 2008.

J. Nickolls, I. Buck, M. Garland and K. Skadron,“Scalable parallel programming with CUDA”, Presentation by Christian Hansen Article Published in ACM Queue, March, page 2, 2008.

S.Tomov, R.Nath, H. Ltaief and J. Dongarra,“Dense linear algebra solvers for multicore with GPU accelerators”, Proc. of IPDPS'10, Atlanta, GA, January 15, pp. 1—2. 2010.

Downloads

Published

2021-12-10

How to Cite

Gichunts, E. E. . (2021). Benchmarking of GPU NVIDIA CUDA, CUBLAS and MAGMA Libraries Based on Matrix Multiplication Problem. Mathematical Problems of Computer Science, 42, 121–126. Retrieved from http://mpcs.sci.am/index.php/mpcs/article/view/223

Issue

Vol. 42 (2014): Mathematical Problems of Computer Science

Section

Articles

License

Creative Commons License

This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.