Energy-aware data prefetching for multi-speed disks

Woo Son Seung, Mahmut Kandemir

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

35 Citations (Scopus)

Abstract

Power consumption of disk based storage systems is becoming an increasingly pressing issue for both commercial and scientific application domains. Prior work proposed several hardware based approaches to reducing disk power consumption by making use of techniques such as spinning down idle disks and rotating them at lower speeds than the maximum speed possible. While such techniques are certainly very important, it is also critical to consider the influence the software can exercise in shaping the power consumption behavior of disk-intensive application programs. Motivated by this, the main goal of this work is to study whether an optimizing compiler can be used for increasing the power benefits that could be obtained from multi-speed disks. Specifically, we propose and experimentally evaluate a compiler-directed energy-aware data prefetching scheme for scientific applications that process disk-resident data sets. This scheme automatically determines the prefetch distance for all disk access instructions, the disk speeds to be employed, and the associated disk layouts (striping parameters) in a unified setting. We implemented the proposed approach within an optimizing compiler framework and conducted experiments with several disk-intensive applications. Our experimental evaluation shows that the proposed approach brings significant reductions in disk energy consumption over a state-of-the-art software-based I/O prefetching mechanism that does not take into account energy consumption explicitly. Our results also show that the energy-aware prefetching scheme does not bring any extra performance penalties and the energy reductions achieved are consistent across a wide spectrum of values of the simulation parameters.

Original languageEnglish (US)
Title of host publicationProceedings of the 3rd Conference on Computing Frontiers 2006, CF '06
Pages105-114
Number of pages10
DOIs
StatePublished - Dec 1 2006
Event3rd Conference on Computing Frontiers 2006, CF '06 - Ischia, Italy
Duration: May 3 2006May 5 2006

Publication series

NameProceedings of the 3rd Conference on Computing Frontiers 2006, CF '06
Volume2006

Other

Other3rd Conference on Computing Frontiers 2006, CF '06
CountryItaly
CityIschia
Period5/3/065/5/06

Fingerprint

Electric power utilization
Energy utilization
Application programs
Hardware
Experiments

All Science Journal Classification (ASJC) codes

  • Computer Science(all)

Cite this

Seung, W. S., & Kandemir, M. (2006). Energy-aware data prefetching for multi-speed disks. In Proceedings of the 3rd Conference on Computing Frontiers 2006, CF '06 (pp. 105-114). (Proceedings of the 3rd Conference on Computing Frontiers 2006, CF '06; Vol. 2006). https://doi.org/10.1145/1128022.1128038
Seung, Woo Son ; Kandemir, Mahmut. / Energy-aware data prefetching for multi-speed disks. Proceedings of the 3rd Conference on Computing Frontiers 2006, CF '06. 2006. pp. 105-114 (Proceedings of the 3rd Conference on Computing Frontiers 2006, CF '06).
@inproceedings{20398839fd8c48a596f29a72adaf80b1,
title = "Energy-aware data prefetching for multi-speed disks",
abstract = "Power consumption of disk based storage systems is becoming an increasingly pressing issue for both commercial and scientific application domains. Prior work proposed several hardware based approaches to reducing disk power consumption by making use of techniques such as spinning down idle disks and rotating them at lower speeds than the maximum speed possible. While such techniques are certainly very important, it is also critical to consider the influence the software can exercise in shaping the power consumption behavior of disk-intensive application programs. Motivated by this, the main goal of this work is to study whether an optimizing compiler can be used for increasing the power benefits that could be obtained from multi-speed disks. Specifically, we propose and experimentally evaluate a compiler-directed energy-aware data prefetching scheme for scientific applications that process disk-resident data sets. This scheme automatically determines the prefetch distance for all disk access instructions, the disk speeds to be employed, and the associated disk layouts (striping parameters) in a unified setting. We implemented the proposed approach within an optimizing compiler framework and conducted experiments with several disk-intensive applications. Our experimental evaluation shows that the proposed approach brings significant reductions in disk energy consumption over a state-of-the-art software-based I/O prefetching mechanism that does not take into account energy consumption explicitly. Our results also show that the energy-aware prefetching scheme does not bring any extra performance penalties and the energy reductions achieved are consistent across a wide spectrum of values of the simulation parameters.",
author = "Seung, {Woo Son} and Mahmut Kandemir",
year = "2006",
month = "12",
day = "1",
doi = "10.1145/1128022.1128038",
language = "English (US)",
isbn = "1595933026",
series = "Proceedings of the 3rd Conference on Computing Frontiers 2006, CF '06",
pages = "105--114",
booktitle = "Proceedings of the 3rd Conference on Computing Frontiers 2006, CF '06",

}

Seung, WS & Kandemir, M 2006, Energy-aware data prefetching for multi-speed disks. in Proceedings of the 3rd Conference on Computing Frontiers 2006, CF '06. Proceedings of the 3rd Conference on Computing Frontiers 2006, CF '06, vol. 2006, pp. 105-114, 3rd Conference on Computing Frontiers 2006, CF '06, Ischia, Italy, 5/3/06. https://doi.org/10.1145/1128022.1128038

Energy-aware data prefetching for multi-speed disks. / Seung, Woo Son; Kandemir, Mahmut.

Proceedings of the 3rd Conference on Computing Frontiers 2006, CF '06. 2006. p. 105-114 (Proceedings of the 3rd Conference on Computing Frontiers 2006, CF '06; Vol. 2006).

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

TY - GEN

T1 - Energy-aware data prefetching for multi-speed disks

AU - Seung, Woo Son

AU - Kandemir, Mahmut

PY - 2006/12/1

Y1 - 2006/12/1

N2 - Power consumption of disk based storage systems is becoming an increasingly pressing issue for both commercial and scientific application domains. Prior work proposed several hardware based approaches to reducing disk power consumption by making use of techniques such as spinning down idle disks and rotating them at lower speeds than the maximum speed possible. While such techniques are certainly very important, it is also critical to consider the influence the software can exercise in shaping the power consumption behavior of disk-intensive application programs. Motivated by this, the main goal of this work is to study whether an optimizing compiler can be used for increasing the power benefits that could be obtained from multi-speed disks. Specifically, we propose and experimentally evaluate a compiler-directed energy-aware data prefetching scheme for scientific applications that process disk-resident data sets. This scheme automatically determines the prefetch distance for all disk access instructions, the disk speeds to be employed, and the associated disk layouts (striping parameters) in a unified setting. We implemented the proposed approach within an optimizing compiler framework and conducted experiments with several disk-intensive applications. Our experimental evaluation shows that the proposed approach brings significant reductions in disk energy consumption over a state-of-the-art software-based I/O prefetching mechanism that does not take into account energy consumption explicitly. Our results also show that the energy-aware prefetching scheme does not bring any extra performance penalties and the energy reductions achieved are consistent across a wide spectrum of values of the simulation parameters.

AB - Power consumption of disk based storage systems is becoming an increasingly pressing issue for both commercial and scientific application domains. Prior work proposed several hardware based approaches to reducing disk power consumption by making use of techniques such as spinning down idle disks and rotating them at lower speeds than the maximum speed possible. While such techniques are certainly very important, it is also critical to consider the influence the software can exercise in shaping the power consumption behavior of disk-intensive application programs. Motivated by this, the main goal of this work is to study whether an optimizing compiler can be used for increasing the power benefits that could be obtained from multi-speed disks. Specifically, we propose and experimentally evaluate a compiler-directed energy-aware data prefetching scheme for scientific applications that process disk-resident data sets. This scheme automatically determines the prefetch distance for all disk access instructions, the disk speeds to be employed, and the associated disk layouts (striping parameters) in a unified setting. We implemented the proposed approach within an optimizing compiler framework and conducted experiments with several disk-intensive applications. Our experimental evaluation shows that the proposed approach brings significant reductions in disk energy consumption over a state-of-the-art software-based I/O prefetching mechanism that does not take into account energy consumption explicitly. Our results also show that the energy-aware prefetching scheme does not bring any extra performance penalties and the energy reductions achieved are consistent across a wide spectrum of values of the simulation parameters.

UR - http://www.scopus.com/inward/record.url?scp=34247366053&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=34247366053&partnerID=8YFLogxK

U2 - 10.1145/1128022.1128038

DO - 10.1145/1128022.1128038

M3 - Conference contribution

SN - 1595933026

SN - 9781595933027

T3 - Proceedings of the 3rd Conference on Computing Frontiers 2006, CF '06

SP - 105

EP - 114

BT - Proceedings of the 3rd Conference on Computing Frontiers 2006, CF '06

ER -

Seung WS, Kandemir M. Energy-aware data prefetching for multi-speed disks. In Proceedings of the 3rd Conference on Computing Frontiers 2006, CF '06. 2006. p. 105-114. (Proceedings of the 3rd Conference on Computing Frontiers 2006, CF '06). https://doi.org/10.1145/1128022.1128038