GreenLA: Green Linear Algebra Software for GPU-accelerated Heterogeneous Computing

Jieyang Chen; Li Tan; Panruo Wu; Dingwen Tao; Hongbo Li; Xin Liang; Sihuan Li; Rong Ge; Laxmi Bhuyan; Zizhong Chen

doi:10.1109/SC.2016.56

GreenLA: Green Linear Algebra Software for GPU-accelerated Heterogeneous Computing

Jieyang Chen, Li Tan, Panruo Wu, Dingwen Tao, Hongbo Li, Xin Liang, Sihuan Li, Rong Ge, Laxmi Bhuyan, Zizhong Chen

Computer Science

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

6 Scopus citations

Abstract

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.

Original language	English
Title of host publication	Proceedings of SC 2016
Subtitle of host publication	The International Conference for High Performance Computing, Networking, Storage and Analysis
Pages	667-677
Number of pages	11
ISBN (Electronic)	9781467388153
DOIs	https://doi.org/10.1109/SC.2016.56
State	Published - Jul 2 2016
Event	2016 International Conference for High Performance Computing, Networking, Storage and Analysis, SC 2016 - Salt Lake City, United States Duration: Nov 13 2016 → Nov 18 2016

Publication series

Name	International Conference for High Performance Computing, Networking, Storage and Analysis, SC
Volume	0
ISSN (Print)	2167-4329
ISSN (Electronic)	2167-4337

Conference

Conference	2016 International Conference for High Performance Computing, Networking, Storage and Analysis, SC 2016
Country/Territory	United States
City	Salt Lake City
Period	11/13/16 → 11/18/16

Bibliographical note

Funding Information:
The authors would like to thank NVIDiA for providing GPU devices used for experiments. This work is partially supported by the NSF grants CCF-1305622, ACI-1305624, CCF-1513201, CCF-1551511, the SZSTI basic research program JCYJ20150630114942313, and the Special Program for Applied Research on Super Computation of the NSFC Guangdong Joint Fund (the second phase).

Publisher Copyright:
© 2016 IEEE.

Keywords

algorithmic slack prediction
CPU
critical path
dense matrix factorizations
DVFS
energy
GPU
performance

ASJC Scopus subject areas

Computer Networks and Communications
Computer Science Applications
Hardware and Architecture
Software

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

Access to Document

10.1109/SC.2016.56

Cite this

Chen, J., Tan, L., Wu, P., Tao, D., Li, H., Liang, X., Li, S., Ge, R., Bhuyan, L., & Chen, Z. (2016). GreenLA: Green Linear Algebra Software for GPU-accelerated Heterogeneous Computing. In Proceedings of SC 2016: The International Conference for High Performance Computing, Networking, Storage and Analysis (pp. 667-677). Article 7877135 (International Conference for High Performance Computing, Networking, Storage and Analysis, SC; Vol. 0). https://doi.org/10.1109/SC.2016.56

@inproceedings{4b2163525cd147aba74833903051c665,

title = "GreenLA: Green Linear Algebra Software for GPU-accelerated Heterogeneous Computing",

abstract = "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.",

keywords = "algorithmic slack prediction, CPU, critical path, dense matrix factorizations, DVFS, energy, GPU, performance",

author = "Jieyang Chen and Li Tan and Panruo Wu and Dingwen Tao and Hongbo Li and Xin Liang and Sihuan Li and Rong Ge and Laxmi Bhuyan and Zizhong Chen",

note = "Funding Information: The authors would like to thank NVIDiA for providing GPU devices used for experiments. This work is partially supported by the NSF grants CCF-1305622, ACI-1305624, CCF-1513201, CCF-1551511, the SZSTI basic research program JCYJ20150630114942313, and the Special Program for Applied Research on Super Computation of the NSFC Guangdong Joint Fund (the second phase). Publisher Copyright: {\textcopyright} 2016 IEEE.; 2016 International Conference for High Performance Computing, Networking, Storage and Analysis, SC 2016 ; Conference date: 13-11-2016 Through 18-11-2016",

year = "2016",

month = jul,

day = "2",

doi = "10.1109/SC.2016.56",

language = "English",

series = "International Conference for High Performance Computing, Networking, Storage and Analysis, SC",

pages = "667--677",

booktitle = "Proceedings of SC 2016",

}

Chen, J, Tan, L, Wu, P, Tao, D, Li, H, Liang, X, Li, S, Ge, R, Bhuyan, L & Chen, Z 2016, GreenLA: Green Linear Algebra Software for GPU-accelerated Heterogeneous Computing. in Proceedings of SC 2016: The International Conference for High Performance Computing, Networking, Storage and Analysis., 7877135, International Conference for High Performance Computing, Networking, Storage and Analysis, SC, vol. 0, pp. 667-677, 2016 International Conference for High Performance Computing, Networking, Storage and Analysis, SC 2016, Salt Lake City, United States, 11/13/16. https://doi.org/10.1109/SC.2016.56

GreenLA: Green Linear Algebra Software for GPU-accelerated Heterogeneous Computing. / Chen, Jieyang; Tan, Li; Wu, Panruo et al.
Proceedings of SC 2016: The International Conference for High Performance Computing, Networking, Storage and Analysis. 2016. p. 667-677 7877135 (International Conference for High Performance Computing, Networking, Storage and Analysis, SC; Vol. 0).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - GreenLA

T2 - 2016 International Conference for High Performance Computing, Networking, Storage and Analysis, SC 2016

AU - Chen, Jieyang

AU - Tan, Li

AU - Wu, Panruo

AU - Tao, Dingwen

AU - Li, Hongbo

AU - Liang, Xin

AU - Li, Sihuan

AU - Ge, Rong

AU - Bhuyan, Laxmi

AU - Chen, Zizhong

N1 - Funding Information: The authors would like to thank NVIDiA for providing GPU devices used for experiments. This work is partially supported by the NSF grants CCF-1305622, ACI-1305624, CCF-1513201, CCF-1551511, the SZSTI basic research program JCYJ20150630114942313, and the Special Program for Applied Research on Super Computation of the NSFC Guangdong Joint Fund (the second phase). Publisher Copyright: © 2016 IEEE.

PY - 2016/7/2

Y1 - 2016/7/2

N2 - 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.

AB - 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.

KW - algorithmic slack prediction

KW - CPU

KW - critical path

KW - dense matrix factorizations

KW - DVFS

KW - energy

KW - GPU

KW - performance

UR - http://www.scopus.com/inward/record.url?scp=85017226199&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85017226199&partnerID=8YFLogxK

U2 - 10.1109/SC.2016.56

DO - 10.1109/SC.2016.56

M3 - Conference contribution

AN - SCOPUS:85017226199

T3 - International Conference for High Performance Computing, Networking, Storage and Analysis, SC

SP - 667

EP - 677

BT - Proceedings of SC 2016

Y2 - 13 November 2016 through 18 November 2016

ER -

Chen J, Tan L, Wu P, Tao D, Li H, Liang X et al. GreenLA: Green Linear Algebra Software for GPU-accelerated Heterogeneous Computing. In Proceedings of SC 2016: The International Conference for High Performance Computing, Networking, Storage and Analysis. 2016. p. 667-677. 7877135. (International Conference for High Performance Computing, Networking, Storage and Analysis, SC). doi: 10.1109/SC.2016.56

GreenLA: Green Linear Algebra Software for GPU-accelerated Heterogeneous Computing

Abstract

Publication series

Conference

Bibliographical note

Keywords

ASJC Scopus subject areas

UN SDGs

Access to Document

Other files and links

Fingerprint

Cite this