On the reliability of drowsy instruction caches

Soong Hyun Shin, Sung Woo Jung, Chu Shik Jhon

Research output: Chapter in Book/Report/Conference proceedingConference contribution

2 Citations (Scopus)

Abstract

As technology scales down, the leakage energy accounts for more portion of total energy in a cache. Applying the Dynamic Voltage Scaling(DVS) to a cache, which is called a drowsy cache, is known as one of the most efficient techniques for reducing leakage energy in a cache. However, it increases the Soft Error Rate(SER) and many researchers began to doubt the reliability of a drowsy cache. In this paper, we show that the instruction cache(I-cache) can adopt the DVS without reliability problems for several reasons. First, an I-cache always stores read-only data, rarely incurring unrecoverable errors. In the I-cache, the soft error can be recovered by re-fetching from the lower level memory. Second, the effect of soft errors on performance is negligible, because the SEE is extremely low. Additional, considerable percentage of soft errors do not harm the performance. In this paper, the evaluation results show that the drowsy I-cache rarely increases unrecoverable errors and negligibly degrades the performance.

Original languageEnglish
Title of host publicationLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Pages445-451
Number of pages7
Volume4186 LNCS
Publication statusPublished - 2006 Dec 8
Event11th Asia-Pacific Conference on Advances in Computer Systems Architecture, ACSAC 2006 - Shanghai, China
Duration: 2006 Sep 62006 Sep 8

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume4186 LNCS
ISSN (Print)03029743
ISSN (Electronic)16113349

Other

Other11th Asia-Pacific Conference on Advances in Computer Systems Architecture, ACSAC 2006
CountryChina
CityShanghai
Period06/9/606/9/8

Fingerprint

Cache
Research Personnel
Technology
Soft Error
Dynamic Voltage Scaling
Leakage
Energy
Data storage equipment
Error Rate
Percentage
Evaluation
Voltage scaling

ASJC Scopus subject areas

  • Biochemistry, Genetics and Molecular Biology(all)
  • Computer Science(all)
  • Theoretical Computer Science

Cite this

Shin, S. H., Jung, S. W., & Jhon, C. S. (2006). On the reliability of drowsy instruction caches. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 4186 LNCS, pp. 445-451). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 4186 LNCS).

On the reliability of drowsy instruction caches. / Shin, Soong Hyun; Jung, Sung Woo; Jhon, Chu Shik.

Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 4186 LNCS 2006. p. 445-451 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 4186 LNCS).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Shin, SH, Jung, SW & Jhon, CS 2006, On the reliability of drowsy instruction caches. in Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). vol. 4186 LNCS, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 4186 LNCS, pp. 445-451, 11th Asia-Pacific Conference on Advances in Computer Systems Architecture, ACSAC 2006, Shanghai, China, 06/9/6.
Shin SH, Jung SW, Jhon CS. On the reliability of drowsy instruction caches. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 4186 LNCS. 2006. p. 445-451. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
Shin, Soong Hyun ; Jung, Sung Woo ; Jhon, Chu Shik. / On the reliability of drowsy instruction caches. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 4186 LNCS 2006. pp. 445-451 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{dfcd684f843449c7b19cc37243e0048e,
title = "On the reliability of drowsy instruction caches",
abstract = "As technology scales down, the leakage energy accounts for more portion of total energy in a cache. Applying the Dynamic Voltage Scaling(DVS) to a cache, which is called a drowsy cache, is known as one of the most efficient techniques for reducing leakage energy in a cache. However, it increases the Soft Error Rate(SER) and many researchers began to doubt the reliability of a drowsy cache. In this paper, we show that the instruction cache(I-cache) can adopt the DVS without reliability problems for several reasons. First, an I-cache always stores read-only data, rarely incurring unrecoverable errors. In the I-cache, the soft error can be recovered by re-fetching from the lower level memory. Second, the effect of soft errors on performance is negligible, because the SEE is extremely low. Additional, considerable percentage of soft errors do not harm the performance. In this paper, the evaluation results show that the drowsy I-cache rarely increases unrecoverable errors and negligibly degrades the performance.",
author = "Shin, {Soong Hyun} and Jung, {Sung Woo} and Jhon, {Chu Shik}",
year = "2006",
month = "12",
day = "8",
language = "English",
isbn = "3540400567",
volume = "4186 LNCS",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
pages = "445--451",
booktitle = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

}

TY - GEN

T1 - On the reliability of drowsy instruction caches

AU - Shin, Soong Hyun

AU - Jung, Sung Woo

AU - Jhon, Chu Shik

PY - 2006/12/8

Y1 - 2006/12/8

N2 - As technology scales down, the leakage energy accounts for more portion of total energy in a cache. Applying the Dynamic Voltage Scaling(DVS) to a cache, which is called a drowsy cache, is known as one of the most efficient techniques for reducing leakage energy in a cache. However, it increases the Soft Error Rate(SER) and many researchers began to doubt the reliability of a drowsy cache. In this paper, we show that the instruction cache(I-cache) can adopt the DVS without reliability problems for several reasons. First, an I-cache always stores read-only data, rarely incurring unrecoverable errors. In the I-cache, the soft error can be recovered by re-fetching from the lower level memory. Second, the effect of soft errors on performance is negligible, because the SEE is extremely low. Additional, considerable percentage of soft errors do not harm the performance. In this paper, the evaluation results show that the drowsy I-cache rarely increases unrecoverable errors and negligibly degrades the performance.

AB - As technology scales down, the leakage energy accounts for more portion of total energy in a cache. Applying the Dynamic Voltage Scaling(DVS) to a cache, which is called a drowsy cache, is known as one of the most efficient techniques for reducing leakage energy in a cache. However, it increases the Soft Error Rate(SER) and many researchers began to doubt the reliability of a drowsy cache. In this paper, we show that the instruction cache(I-cache) can adopt the DVS without reliability problems for several reasons. First, an I-cache always stores read-only data, rarely incurring unrecoverable errors. In the I-cache, the soft error can be recovered by re-fetching from the lower level memory. Second, the effect of soft errors on performance is negligible, because the SEE is extremely low. Additional, considerable percentage of soft errors do not harm the performance. In this paper, the evaluation results show that the drowsy I-cache rarely increases unrecoverable errors and negligibly degrades the performance.

UR - http://www.scopus.com/inward/record.url?scp=33845215138&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=33845215138&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:33845215138

SN - 3540400567

SN - 9783540400561

VL - 4186 LNCS

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 445

EP - 451

BT - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

ER -