Performance linked dynamic cache tuning: A static energy reduction approach in tiled CMPs

被引:6
作者
Chakraborty, Shounak [1 ]
Kapoor, Hemangee K. [1 ]
机构
[1] Indian Inst Technol Guwahati, Dept Comp Sci & Engn, Gauhati 781039, Assam, India
关键词
Power optimization; Last level cache; Chip multiprocessor; Dynamic power; Leakage power; Performance; Cache power;
D O I
10.1016/j.micpro.2017.06.012
中图分类号
TP3 [计算技术、计算机技术];
学科分类号
0812 ;
摘要
Advancement in semiconductor technology increases power density in recent Chip Multi-Processors (CMPs) which significantly increases the leakage energy consumptions of on-chip Last Level Caches (LLCs). Performance linked dynamic tuning in LLC size is a promising option for reducing the cache leakage. This paper reduces static power consumption by dynamically shutting down or turning on cache banks based upon system performance and cache bank usage statistics. Shutting down of a cache bank remaps its future requests to another active bank, called as target bank. The proposed method is evaluated on three different implementation policies, viz (1) The system can decide to shutdown or turn-on some cache banks periodically throughout the process execution. (2) The system allows to shutdown banks initially and once the bank restarting initiates, no more shutdown is permitted further. (3) This policy resizes cache like first policy with some predefined time slices, in which cache cannot be resized. For a 4MB 4 way set associative L2 cache, experimental analysis shows 66% reduction in static energy with 29% gain in Energy Delay Product (EDP) for first strategy; for the second policy, static power is reduced by 59% with 27% savings in EDP. Finally, last policy saves 65% in static power and 30% in EDP with minimal performance penalty. (C) 2017 Elsevier B.V. All rights reserved.
引用
收藏
页码:221 / 235
页数:15
相关论文
共 23 条
[1]   GARNET: A Detailed On-Chip Network Model inside a Full-System Simulator [J].
Agarwal, Niket ;
Krishna, Tushar ;
Peh, Li-Shiuan ;
Jha, Niraj K. .
ISPASS 2009: IEEE INTERNATIONAL SYMPOSIUM ON PERFORMANCE ANALYSIS OF SYSTEMS AND SOFTWARE, 2009, :33-42
[2]   Way adaptable D-NUCA caches [J].
Bardine A. ;
Comparetti M. ;
Foglia P. ;
Gabrielli G. ;
Prete C.A. .
International Journal of High Performance Systems Architecture, 2010, 2 (3-4) :215-228
[3]  
Bardine A., P 2007 WORKSHOP MEMO, p2007, pp. 105, DOI [10.1145/1327171.1327184, DOI 10.1145/1327171.1327184]
[4]  
Bienia C, 2008, TR81108 PRINC U
[5]  
Dani A. M., 2011, Proceedings of the 2011 Second Workshop on Architecture and Multi-Core Applications (WAMCA 2011), P12, DOI 10.1109/WAMCA.2011.14
[6]   A Framework for Block Placement, Migration, and Fast Searching in Tiled-DNUCA Architecture [J].
Das, Shirshendu ;
Kapoor, Hemangee K. .
ACM TRANSACTIONS ON DESIGN AUTOMATION OF ELECTRONIC SYSTEMS, 2016, 22 (01)
[7]  
Fitzgerald B., 2013, INT GREEN COMP C IGC, P1, DOI DOI 10.1109/IGCC.2013.6604475
[8]   A workload independent energy reduction strategy for D-NUCA caches [J].
Foglia, Pierfrancesco ;
Comparetti, Manuel .
JOURNAL OF SUPERCOMPUTING, 2014, 68 (01) :157-182
[9]  
Kapoor H.K., 2015, 30 ACM SIGAPP S APPL
[10]  
Kim NS, 2003, COMPUTER, V36, P68, DOI 10.1109/MC.2003.1250885