DocumentCode
39919
Title
Many-Core Token-Based Adaptive Power Gating
Author
Kahng, Andrew ; Seokhyeong Kang ; Rosing, Tajana Simunic ; Strong, Ray
Author_Institution
Univ. of California, San Diego, La Jolla, CA, USA
Volume
32
Issue
8
fYear
2013
fDate
Aug. 2013
Firstpage
1288
Lastpage
1292
Abstract
Among power dissipation components, leakage power has become more dominant with each successive technology node. Leakage energy waste can be reduced by power gating. In this paper, we extend token-based adaptive power gating (TAP), a technique to power gate an actively executing core during memory accesses, to many-core Chip Multi-Processors (CMPs). TAP works by tracking every system memory request and its estimated time of arrival so that a core may power gate itself without performance or energy loss. Previous work on TAP [11] shows several benefits compared to earlier state-of-the-art techniques [10], including zero performance hit and 2.58 times average energy savings for out-of-order cores. We show that TAP can adapt to increasing memory contention by increasing power-gated time by 3.69 times compared to a low memory-pressure case. We also scale TAP to many-core architectures with a distributed wake-up controller that is capable of supporting staggered wake-ups and able to power gate each core for 99.07% of the time, achieved by a non-scalable centralized scheme.
Keywords
cache storage; microprocessor chips; multiprocessing systems; CMP; TAP; distributed wake-up controller; energy savings; leakage energy waste; leakage power; many-core chip multiprocessors; many-core token; memory access; nonscalable centralized scheme; power dissipation component; token-based adaptive power gating; Benchmark testing; Delays; Instruction sets; Logic gates; Multicore processing; Noise; Adaptive power gating; energy savings; low power design; many-core architecture;
fLanguage
English
Journal_Title
Computer-Aided Design of Integrated Circuits and Systems, IEEE Transactions on
Publisher
ieee
ISSN
0278-0070
Type
jour
DOI
10.1109/TCAD.2013.2257923
Filename
6559076
Link To Document