An effective matrix compression method for GPU-accelerated thermal analysis

Lih Yih Chiou; Liang Ying Lu; Chieh Yu Lin

doi:10.1109/VLSI-DAT.2015.7114505

An effective matrix compression method for GPU-accelerated thermal analysis

Lih Yih Chiou, Liang Ying Lu, Chieh Yu Lin

電機工程學系

研究成果: Conference contribution

摘要

Three-dimensional integrated circuits are expected to face increasingly severe thermal challenges and cost issues as the number of stacked ICs increases. Thermal analysis for 3D ICs is urgently required to assist system designers at the early phase of design to identify hot zones. Most thermal analyses obtain detailed temperature distribution by large matrix operations, and hence reduce analysis performance. Accordingly, we propose a compressed and combined sparse row (CCSR) matrix format to be used in the proposed effective matrix compression (EMC) method for matrix multiplication on GPU. The experimental results show EMC using CCSR is on average 44.93 times faster than matrix multiplication without any special compression format and on average at least 3.09 times faster than other compression formats.

原文	English
主出版物標題	2015 International Symposium on VLSI Design, Automation and Test, VLSI-DAT 2015
發行者	Institute of Electrical and Electronics Engineers Inc.
ISBN（電子）	9781479962754
DOIs	https://doi.org/10.1109/VLSI-DAT.2015.7114505
出版狀態	Published - 2015 5月 28
事件	2015 International Symposium on VLSI Design, Automation and Test, VLSI-DAT 2015 - Hsinchu, Taiwan 持續時間: 2015 4月 27 → 2015 4月 29

出版系列

名字	2015 International Symposium on VLSI Design, Automation and Test, VLSI-DAT 2015

Other

Other	2015 International Symposium on VLSI Design, Automation and Test, VLSI-DAT 2015
國家/地區	Taiwan
城市	Hsinchu
期間	15-04-27 → 15-04-29

All Science Journal Classification (ASJC) codes

硬體和架構
電氣與電子工程

存取文件

10.1109/VLSI-DAT.2015.7114505

其它文件與鏈接

引用此

Chiou, L. Y., Lu, L. Y., & Lin, C. Y. (2015). An effective matrix compression method for GPU-accelerated thermal analysis. 於 2015 International Symposium on VLSI Design, Automation and Test, VLSI-DAT 2015 文章 7114505 (2015 International Symposium on VLSI Design, Automation and Test, VLSI-DAT 2015). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/VLSI-DAT.2015.7114505

@inproceedings{26101feafb2e476b8d05cbb0f5c7d799,

title = "An effective matrix compression method for GPU-accelerated thermal analysis",

abstract = "Three-dimensional integrated circuits are expected to face increasingly severe thermal challenges and cost issues as the number of stacked ICs increases. Thermal analysis for 3D ICs is urgently required to assist system designers at the early phase of design to identify hot zones. Most thermal analyses obtain detailed temperature distribution by large matrix operations, and hence reduce analysis performance. Accordingly, we propose a compressed and combined sparse row (CCSR) matrix format to be used in the proposed effective matrix compression (EMC) method for matrix multiplication on GPU. The experimental results show EMC using CCSR is on average 44.93 times faster than matrix multiplication without any special compression format and on average at least 3.09 times faster than other compression formats.",

author = "Chiou, {Lih Yih} and Lu, {Liang Ying} and Lin, {Chieh Yu}",

note = "Publisher Copyright: {\textcopyright} 2015 IEEE.; 2015 International Symposium on VLSI Design, Automation and Test, VLSI-DAT 2015 ; Conference date: 27-04-2015 Through 29-04-2015",

year = "2015",

month = may,

day = "28",

doi = "10.1109/VLSI-DAT.2015.7114505",

language = "English",

series = "2015 International Symposium on VLSI Design, Automation and Test, VLSI-DAT 2015",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

booktitle = "2015 International Symposium on VLSI Design, Automation and Test, VLSI-DAT 2015",

address = "United States",

}

Chiou, LY, Lu, LY & Lin, CY 2015, An effective matrix compression method for GPU-accelerated thermal analysis. 於 2015 International Symposium on VLSI Design, Automation and Test, VLSI-DAT 2015., 7114505, 2015 International Symposium on VLSI Design, Automation and Test, VLSI-DAT 2015, Institute of Electrical and Electronics Engineers Inc., 2015 International Symposium on VLSI Design, Automation and Test, VLSI-DAT 2015, Hsinchu, Taiwan, 15-04-27. https://doi.org/10.1109/VLSI-DAT.2015.7114505

An effective matrix compression method for GPU-accelerated thermal analysis. / Chiou, Lih Yih; Lu, Liang Ying; Lin, Chieh Yu.
2015 International Symposium on VLSI Design, Automation and Test, VLSI-DAT 2015. Institute of Electrical and Electronics Engineers Inc., 2015. 7114505 (2015 International Symposium on VLSI Design, Automation and Test, VLSI-DAT 2015).

研究成果: Conference contribution

TY - GEN

T1 - An effective matrix compression method for GPU-accelerated thermal analysis

AU - Chiou, Lih Yih

AU - Lu, Liang Ying

AU - Lin, Chieh Yu

PY - 2015/5/28

Y1 - 2015/5/28

N2 - Three-dimensional integrated circuits are expected to face increasingly severe thermal challenges and cost issues as the number of stacked ICs increases. Thermal analysis for 3D ICs is urgently required to assist system designers at the early phase of design to identify hot zones. Most thermal analyses obtain detailed temperature distribution by large matrix operations, and hence reduce analysis performance. Accordingly, we propose a compressed and combined sparse row (CCSR) matrix format to be used in the proposed effective matrix compression (EMC) method for matrix multiplication on GPU. The experimental results show EMC using CCSR is on average 44.93 times faster than matrix multiplication without any special compression format and on average at least 3.09 times faster than other compression formats.

AB - Three-dimensional integrated circuits are expected to face increasingly severe thermal challenges and cost issues as the number of stacked ICs increases. Thermal analysis for 3D ICs is urgently required to assist system designers at the early phase of design to identify hot zones. Most thermal analyses obtain detailed temperature distribution by large matrix operations, and hence reduce analysis performance. Accordingly, we propose a compressed and combined sparse row (CCSR) matrix format to be used in the proposed effective matrix compression (EMC) method for matrix multiplication on GPU. The experimental results show EMC using CCSR is on average 44.93 times faster than matrix multiplication without any special compression format and on average at least 3.09 times faster than other compression formats.

UR - http://www.scopus.com/inward/record.url?scp=84936984417&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84936984417&partnerID=8YFLogxK

U2 - 10.1109/VLSI-DAT.2015.7114505

DO - 10.1109/VLSI-DAT.2015.7114505

M3 - Conference contribution

AN - SCOPUS:84936984417

T3 - 2015 International Symposium on VLSI Design, Automation and Test, VLSI-DAT 2015

BT - 2015 International Symposium on VLSI Design, Automation and Test, VLSI-DAT 2015

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 2015 International Symposium on VLSI Design, Automation and Test, VLSI-DAT 2015

Y2 - 27 April 2015 through 29 April 2015

ER -

Chiou LY, Lu LY, Lin CY. An effective matrix compression method for GPU-accelerated thermal analysis. 於 2015 International Symposium on VLSI Design, Automation and Test, VLSI-DAT 2015. Institute of Electrical and Electronics Engineers Inc. 2015. 7114505. (2015 International Symposium on VLSI Design, Automation and Test, VLSI-DAT 2015). doi: 10.1109/VLSI-DAT.2015.7114505

An effective matrix compression method for GPU-accelerated thermal analysis

摘要

出版系列

Other

All Science Journal Classification (ASJC) codes

存取文件

其它文件與鏈接

指紋

引用此