A survey of system level power management schemes in the dark-silicon era for many-core architectures

Research output: Contribution to JournalArticle

Abstract

Power consumption in Complementary Metal Oxide Semiconductor (CMOS) technology has escalated to a point that only a fractional part of many-core chips can be powered-on at a time. Fortunately, this fraction can be increased at the expense of performance through the dark-silicon solution. However, with many-core integration set to be heading towards its thousands, power consumption and temperature increases per time, meaning the number of active nodes must be reduced drastically. Therefore, optimized techniques are demanded for continuous advancement in technology. Existing efforts try to overcome this challenge by activating nodes from different parts of the chip at the expense of communication latency. Other efforts on the other hand employ run-time power management techniques to manage the power performance of the cores trading-off performance for power. We found out that, for a significant amount of power to saved and high temperature to be avoided, focus should be on reducing the power consumption of all the on-chip components. Especially, the memory hierarchy and the interconnect. Power consumption can be minimized by, reducing the size of high leakage power dissipating elements, turning-off idle resources and integrating power saving materials.
Original languageEnglish
JournalEAI Endorsed Transactions on Industrial Networks and Intelligent Systems
Volume5
Issue number15
DOIs
Publication statusPublished - 19 Sep 2018

Fingerprint

Electric power utilization
Silicon
Data storage equipment
Temperature
Power management
Communication
Metals

Keywords

  • Dark-silicon
  • many-core
  • multi-core
  • NoC
  • low power NoC architectures
  • power budgeting

Cite this

@article{9fb38482dc1846a8b7f360364078b056,
title = "A survey of system level power management schemes in the dark-silicon era for many-core architectures",
abstract = "Power consumption in Complementary Metal Oxide Semiconductor (CMOS) technology has escalated to a point that only a fractional part of many-core chips can be powered-on at a time. Fortunately, this fraction can be increased at the expense of performance through the dark-silicon solution. However, with many-core integration set to be heading towards its thousands, power consumption and temperature increases per time, meaning the number of active nodes must be reduced drastically. Therefore, optimized techniques are demanded for continuous advancement in technology. Existing efforts try to overcome this challenge by activating nodes from different parts of the chip at the expense of communication latency. Other efforts on the other hand employ run-time power management techniques to manage the power performance of the cores trading-off performance for power. We found out that, for a significant amount of power to saved and high temperature to be avoided, focus should be on reducing the power consumption of all the on-chip components. Especially, the memory hierarchy and the interconnect. Power consumption can be minimized by, reducing the size of high leakage power dissipating elements, turning-off idle resources and integrating power saving materials.",
keywords = "Dark-silicon, many-core, multi-core, NoC, low power NoC architectures, power budgeting",
author = "Emmanuel Ofori-Attah and Xiaohang Wang and {Opoku Agyeman}, Michael",
year = "2018",
month = "9",
day = "19",
doi = "10.4108/eai.19-9-2018.155569",
language = "English",
volume = "5",
journal = "EAI Endorsed Transactions on Industrial Networks and Intelligent Systems",
issn = "2410-0218",
number = "15",

}

TY - JOUR

T1 - A survey of system level power management schemes in the dark-silicon era for many-core architectures

AU - Ofori-Attah, Emmanuel

AU - Wang, Xiaohang

AU - Opoku Agyeman, Michael

PY - 2018/9/19

Y1 - 2018/9/19

N2 - Power consumption in Complementary Metal Oxide Semiconductor (CMOS) technology has escalated to a point that only a fractional part of many-core chips can be powered-on at a time. Fortunately, this fraction can be increased at the expense of performance through the dark-silicon solution. However, with many-core integration set to be heading towards its thousands, power consumption and temperature increases per time, meaning the number of active nodes must be reduced drastically. Therefore, optimized techniques are demanded for continuous advancement in technology. Existing efforts try to overcome this challenge by activating nodes from different parts of the chip at the expense of communication latency. Other efforts on the other hand employ run-time power management techniques to manage the power performance of the cores trading-off performance for power. We found out that, for a significant amount of power to saved and high temperature to be avoided, focus should be on reducing the power consumption of all the on-chip components. Especially, the memory hierarchy and the interconnect. Power consumption can be minimized by, reducing the size of high leakage power dissipating elements, turning-off idle resources and integrating power saving materials.

AB - Power consumption in Complementary Metal Oxide Semiconductor (CMOS) technology has escalated to a point that only a fractional part of many-core chips can be powered-on at a time. Fortunately, this fraction can be increased at the expense of performance through the dark-silicon solution. However, with many-core integration set to be heading towards its thousands, power consumption and temperature increases per time, meaning the number of active nodes must be reduced drastically. Therefore, optimized techniques are demanded for continuous advancement in technology. Existing efforts try to overcome this challenge by activating nodes from different parts of the chip at the expense of communication latency. Other efforts on the other hand employ run-time power management techniques to manage the power performance of the cores trading-off performance for power. We found out that, for a significant amount of power to saved and high temperature to be avoided, focus should be on reducing the power consumption of all the on-chip components. Especially, the memory hierarchy and the interconnect. Power consumption can be minimized by, reducing the size of high leakage power dissipating elements, turning-off idle resources and integrating power saving materials.

KW - Dark-silicon

KW - many-core

KW - multi-core

KW - NoC

KW - low power NoC architectures

KW - power budgeting

U2 - 10.4108/eai.19-9-2018.155569

DO - 10.4108/eai.19-9-2018.155569

M3 - Article

VL - 5

JO - EAI Endorsed Transactions on Industrial Networks and Intelligent Systems

JF - EAI Endorsed Transactions on Industrial Networks and Intelligent Systems

SN - 2410-0218

IS - 15

ER -