Home // INFOCOMP 2015, The Fifth International Conference on Advanced Communications and Computation // View article

Gauss-Jordan Matrix Inversion Speed-Up using GPUs with the Consideration of Power Consumption

Authors:
Mahmoud Shirazi
Mehdi Kargahi
Farshad Khunjush

Keywords: Matrix Inversion; GPU; Power Consumption

Abstract:
Matrix inversion is an important requirement for many scientific and practical applications. Parallel architectures like Graphics Processing Units (GPUs) can have noticeable impacts on accelerating such computations, however, at the expense of more power consumption. Both computation speed and power become more challenging when the size of matrices gets larger. This paper proposes an accelerated version of the Gauss-Jordan matrix inversion method using GPUs. The experimental results show that the proposed method is faster than a recent baseline method. As an additional property, the configurations have been made in a manner to consume less power.

Pages: 20 to 25

Copyright: Copyright (c) IARIA, 2015

Publication date: June 21, 2015

Published in: conference

ISSN: 2308-3484

ISBN: 978-1-61208-416-9

Location: Brussels, Belgium

Dates: from June 21, 2015 to June 26, 2015