GreenLA: Green Linear Algebra Software for GPU-Accelerated Heterogeneous Computing
While many linear algebra libraries have been developed to optimize their performance, no linear algebra library considers their energy efficiency at the library design time. In this paper, we present GreenLA- A n energy efficient linear algebra software package that leverages linear algebra algorithmic characteristics to maximize energy savings with negligible overhead. GreenLA is (1) energy efficient: It saves up to several times more energy than the best existing energy saving approaches that do not modify library source codes; (2) high performance: Its performance is comparable to the highly optimized linear algebra library MAGMA; and (3) transparent to applications: With the same programming interface, existing MAGMA users do not need to modify their source codes to benefit from GreenLA. Experimental results demonstrate that GreenLA is able to save up to three times more energy than the best existing energy saving approaches while delivering similar performance compared to the state-of-the-art linear algebra library MAGMA.
J. Chen et al., "GreenLA: Green Linear Algebra Software for GPU-Accelerated Heterogeneous Computing," Proceedings of the International Conference for High Performance Computing, Networking, Storage and Analysis (2016, Salt Lake City, UT), pp. 667-677, Association for Computing Machinery (ACM), Jul 2016.
The definitive version is available at https://doi.org/10.1109/SC.2016.56
International Conference for High Performance Computing, Networking, Storage and Analysis, SC '16 (2016: Nov. 13-18, Salt Lake City, UT)
Keywords and Phrases
Algorithmic Slack Prediction; CPU; Critical Path; Dense Matrix Factorizations; DVFS; Energy; GPU; Performance
International Standard Book Number (ISBN)
International Standard Serial Number (ISSN)
Article - Conference proceedings
© 2016 Association for Computing Machinery (ACM), All rights reserved.
02 Jul 2016