Lowering minimum supply voltage for power-efficient cache design by exploiting data redundancy

Dongha Jung, Hokyoon Lee, Seon Wook Kim

Research output: Contribution to journalArticle

2 Citations (Scopus)

Abstract

Dynamic voltage scaling, SRAM reliability, cache replacement policy, VCCMIN Voltage scaling is known to be an efficient way of saving power and energy within a system, and large caches such as LLCs are good candidates for voltage scaling considering their constantly increasing size. However, the VCCMIN problem, in which the lower bound of scalable voltage is limited by process variation, has made it difficult to exploit the benefits of voltage scaling. Lowering VCCMIN incurs multibit faults, which cannot be efficiently resolved by current technologies due to their high complexity and power consumption. We overcame the limitation by exploiting the data redundancy of memory hierarchy. For example, cache coherence states and several layers of cache organization naturally expose the existence of redundancy within cache blocks. If blocks have redundant copies, their VCCMIN can be lowered; although more faults can occur in the blocks, they can be efficiently detected by simple error detection codes and recovered by reloading the redundant copies. Our scheme requires only minor modifications to the existing cache design. We verified our proposal on a cycle accurate simulator with SPLASH-2 and PARSEC benchmark suites and found that the VCCMIN of a 2MB L2 cache can be further lowered by 0.1V in 32nm technology with negligible degradation in performance. As a result, we could achieve 15.6% of reduction in dynamic power and 15.4% of reduction in static power compared to the previous minimum power.

Original languageEnglish
Article number11
JournalACM Transactions on Design Automation of Electronic Systems
Volume21
Issue number1
DOIs
Publication statusPublished - 2015 Nov 1

Fingerprint

Redundancy
Electric potential
Static random access storage
Error detection
Electric power utilization
Simulators
Data storage equipment
Degradation
Voltage scaling

ASJC Scopus subject areas

  • Computer Science Applications
  • Computer Graphics and Computer-Aided Design
  • Electrical and Electronic Engineering

Cite this

Lowering minimum supply voltage for power-efficient cache design by exploiting data redundancy. / Jung, Dongha; Lee, Hokyoon; Kim, Seon Wook.

In: ACM Transactions on Design Automation of Electronic Systems, Vol. 21, No. 1, 11, 01.11.2015.

Research output: Contribution to journalArticle

@article{534a16679bd34729837a63131e6da7a4,
title = "Lowering minimum supply voltage for power-efficient cache design by exploiting data redundancy",
abstract = "Dynamic voltage scaling, SRAM reliability, cache replacement policy, VCCMIN Voltage scaling is known to be an efficient way of saving power and energy within a system, and large caches such as LLCs are good candidates for voltage scaling considering their constantly increasing size. However, the VCCMIN problem, in which the lower bound of scalable voltage is limited by process variation, has made it difficult to exploit the benefits of voltage scaling. Lowering VCCMIN incurs multibit faults, which cannot be efficiently resolved by current technologies due to their high complexity and power consumption. We overcame the limitation by exploiting the data redundancy of memory hierarchy. For example, cache coherence states and several layers of cache organization naturally expose the existence of redundancy within cache blocks. If blocks have redundant copies, their VCCMIN can be lowered; although more faults can occur in the blocks, they can be efficiently detected by simple error detection codes and recovered by reloading the redundant copies. Our scheme requires only minor modifications to the existing cache design. We verified our proposal on a cycle accurate simulator with SPLASH-2 and PARSEC benchmark suites and found that the VCCMIN of a 2MB L2 cache can be further lowered by 0.1V in 32nm technology with negligible degradation in performance. As a result, we could achieve 15.6{\%} of reduction in dynamic power and 15.4{\%} of reduction in static power compared to the previous minimum power.",
author = "Dongha Jung and Hokyoon Lee and Kim, {Seon Wook}",
year = "2015",
month = "11",
day = "1",
doi = "10.1145/2795229",
language = "English",
volume = "21",
journal = "ACM Transactions on Design Automation of Electronic Systems",
issn = "1084-4309",
publisher = "Association for Computing Machinery (ACM)",
number = "1",

}

TY - JOUR

T1 - Lowering minimum supply voltage for power-efficient cache design by exploiting data redundancy

AU - Jung, Dongha

AU - Lee, Hokyoon

AU - Kim, Seon Wook

PY - 2015/11/1

Y1 - 2015/11/1

N2 - Dynamic voltage scaling, SRAM reliability, cache replacement policy, VCCMIN Voltage scaling is known to be an efficient way of saving power and energy within a system, and large caches such as LLCs are good candidates for voltage scaling considering their constantly increasing size. However, the VCCMIN problem, in which the lower bound of scalable voltage is limited by process variation, has made it difficult to exploit the benefits of voltage scaling. Lowering VCCMIN incurs multibit faults, which cannot be efficiently resolved by current technologies due to their high complexity and power consumption. We overcame the limitation by exploiting the data redundancy of memory hierarchy. For example, cache coherence states and several layers of cache organization naturally expose the existence of redundancy within cache blocks. If blocks have redundant copies, their VCCMIN can be lowered; although more faults can occur in the blocks, they can be efficiently detected by simple error detection codes and recovered by reloading the redundant copies. Our scheme requires only minor modifications to the existing cache design. We verified our proposal on a cycle accurate simulator with SPLASH-2 and PARSEC benchmark suites and found that the VCCMIN of a 2MB L2 cache can be further lowered by 0.1V in 32nm technology with negligible degradation in performance. As a result, we could achieve 15.6% of reduction in dynamic power and 15.4% of reduction in static power compared to the previous minimum power.

AB - Dynamic voltage scaling, SRAM reliability, cache replacement policy, VCCMIN Voltage scaling is known to be an efficient way of saving power and energy within a system, and large caches such as LLCs are good candidates for voltage scaling considering their constantly increasing size. However, the VCCMIN problem, in which the lower bound of scalable voltage is limited by process variation, has made it difficult to exploit the benefits of voltage scaling. Lowering VCCMIN incurs multibit faults, which cannot be efficiently resolved by current technologies due to their high complexity and power consumption. We overcame the limitation by exploiting the data redundancy of memory hierarchy. For example, cache coherence states and several layers of cache organization naturally expose the existence of redundancy within cache blocks. If blocks have redundant copies, their VCCMIN can be lowered; although more faults can occur in the blocks, they can be efficiently detected by simple error detection codes and recovered by reloading the redundant copies. Our scheme requires only minor modifications to the existing cache design. We verified our proposal on a cycle accurate simulator with SPLASH-2 and PARSEC benchmark suites and found that the VCCMIN of a 2MB L2 cache can be further lowered by 0.1V in 32nm technology with negligible degradation in performance. As a result, we could achieve 15.6% of reduction in dynamic power and 15.4% of reduction in static power compared to the previous minimum power.

UR - http://www.scopus.com/inward/record.url?scp=84952011765&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84952011765&partnerID=8YFLogxK

U2 - 10.1145/2795229

DO - 10.1145/2795229

M3 - Article

AN - SCOPUS:84952011765

VL - 21

JO - ACM Transactions on Design Automation of Electronic Systems

JF - ACM Transactions on Design Automation of Electronic Systems

SN - 1084-4309

IS - 1

M1 - 11

ER -