Reducing the impact of intra-core process variability with criticality-based resource allocation and prefetching

Bogdan F. Romanescu, Michael E. Bauer, Sule Ozev, Daniel J. Sorin

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

12 Citations (Scopus)

Abstract

We develop architectural techniques for mitigating the impact of process variability. Our techniques hide the performance effects of slow components-including registers, functional units, and L1I and L1D cache frames-without slowing the clock frequency or pessimistically assuming that all components are slow. Using ideas previously developed for other purposes-criticality-based allocation of resources, prefetching, and prefetch buffering-we allow design engineers to aggressively set the clock frequency without worrying about the subset of components that cannot meet this frequency. Our techniques outperform speed binning, because clock frequency benefits outweigh slight losses in IPC.

Original languageEnglish (US)
Title of host publicationConference on Computing Frontiers - Proceedings of the 2008 Conference on Computing Frontiers, CF'08
Pages129-138
Number of pages10
DOIs
StatePublished - 2008
Externally publishedYes
Event2008 Conference on Computing Frontiers, CF'08 - Ischia, Italy
Duration: May 5 2008May 7 2008

Other

Other2008 Conference on Computing Frontiers, CF'08
CountryItaly
CityIschia
Period5/5/085/7/08

Fingerprint

Resource allocation
Clocks
Engineers

Keywords

  • Microarchitecture
  • Process variability

ASJC Scopus subject areas

  • Computer Science Applications
  • Hardware and Architecture
  • Software
  • Electrical and Electronic Engineering

Cite this

Romanescu, B. F., Bauer, M. E., Ozev, S., & Sorin, D. J. (2008). Reducing the impact of intra-core process variability with criticality-based resource allocation and prefetching. In Conference on Computing Frontiers - Proceedings of the 2008 Conference on Computing Frontiers, CF'08 (pp. 129-138) https://doi.org/10.1145/1366230.1366257

Reducing the impact of intra-core process variability with criticality-based resource allocation and prefetching. / Romanescu, Bogdan F.; Bauer, Michael E.; Ozev, Sule; Sorin, Daniel J.

Conference on Computing Frontiers - Proceedings of the 2008 Conference on Computing Frontiers, CF'08. 2008. p. 129-138.

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

Romanescu, BF, Bauer, ME, Ozev, S & Sorin, DJ 2008, Reducing the impact of intra-core process variability with criticality-based resource allocation and prefetching. in Conference on Computing Frontiers - Proceedings of the 2008 Conference on Computing Frontiers, CF'08. pp. 129-138, 2008 Conference on Computing Frontiers, CF'08, Ischia, Italy, 5/5/08. https://doi.org/10.1145/1366230.1366257
Romanescu BF, Bauer ME, Ozev S, Sorin DJ. Reducing the impact of intra-core process variability with criticality-based resource allocation and prefetching. In Conference on Computing Frontiers - Proceedings of the 2008 Conference on Computing Frontiers, CF'08. 2008. p. 129-138 https://doi.org/10.1145/1366230.1366257
Romanescu, Bogdan F. ; Bauer, Michael E. ; Ozev, Sule ; Sorin, Daniel J. / Reducing the impact of intra-core process variability with criticality-based resource allocation and prefetching. Conference on Computing Frontiers - Proceedings of the 2008 Conference on Computing Frontiers, CF'08. 2008. pp. 129-138
@inproceedings{65ba8ba1f65a4204a1a9c2b384434bc2,
title = "Reducing the impact of intra-core process variability with criticality-based resource allocation and prefetching",
abstract = "We develop architectural techniques for mitigating the impact of process variability. Our techniques hide the performance effects of slow components-including registers, functional units, and L1I and L1D cache frames-without slowing the clock frequency or pessimistically assuming that all components are slow. Using ideas previously developed for other purposes-criticality-based allocation of resources, prefetching, and prefetch buffering-we allow design engineers to aggressively set the clock frequency without worrying about the subset of components that cannot meet this frequency. Our techniques outperform speed binning, because clock frequency benefits outweigh slight losses in IPC.",
keywords = "Microarchitecture, Process variability",
author = "Romanescu, {Bogdan F.} and Bauer, {Michael E.} and Sule Ozev and Sorin, {Daniel J.}",
year = "2008",
doi = "10.1145/1366230.1366257",
language = "English (US)",
isbn = "9781605580777",
pages = "129--138",
booktitle = "Conference on Computing Frontiers - Proceedings of the 2008 Conference on Computing Frontiers, CF'08",

}

TY - GEN

T1 - Reducing the impact of intra-core process variability with criticality-based resource allocation and prefetching

AU - Romanescu, Bogdan F.

AU - Bauer, Michael E.

AU - Ozev, Sule

AU - Sorin, Daniel J.

PY - 2008

Y1 - 2008

N2 - We develop architectural techniques for mitigating the impact of process variability. Our techniques hide the performance effects of slow components-including registers, functional units, and L1I and L1D cache frames-without slowing the clock frequency or pessimistically assuming that all components are slow. Using ideas previously developed for other purposes-criticality-based allocation of resources, prefetching, and prefetch buffering-we allow design engineers to aggressively set the clock frequency without worrying about the subset of components that cannot meet this frequency. Our techniques outperform speed binning, because clock frequency benefits outweigh slight losses in IPC.

AB - We develop architectural techniques for mitigating the impact of process variability. Our techniques hide the performance effects of slow components-including registers, functional units, and L1I and L1D cache frames-without slowing the clock frequency or pessimistically assuming that all components are slow. Using ideas previously developed for other purposes-criticality-based allocation of resources, prefetching, and prefetch buffering-we allow design engineers to aggressively set the clock frequency without worrying about the subset of components that cannot meet this frequency. Our techniques outperform speed binning, because clock frequency benefits outweigh slight losses in IPC.

KW - Microarchitecture

KW - Process variability

UR - http://www.scopus.com/inward/record.url?scp=56749160835&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=56749160835&partnerID=8YFLogxK

U2 - 10.1145/1366230.1366257

DO - 10.1145/1366230.1366257

M3 - Conference contribution

SN - 9781605580777

SP - 129

EP - 138

BT - Conference on Computing Frontiers - Proceedings of the 2008 Conference on Computing Frontiers, CF'08

ER -