Advanced search
1 file | 1.81 MB

A counter architecture for online DVFS profitability estimation

Stijn Eyerman (UGent) and Lieven Eeckhout (UGent)
(2010) IEEE TRANSACTIONS ON COMPUTERS. 59(11). p.1576-1583
Author
Organization
Project
HPC-UGent: the central High Performance Computing infrastructure of Ghent University
Abstract
Dynamic voltage and frequency scaling (DVFS) is a well-known and effective technique for reducing power consumption in modern microprocessors. An important concern though when applying DVFS for online energy and power optimizations is to estimate its profitability in terms of performance and energy. Current DVFS profitability estimation approaches however lack accuracy or incur runtime performance and/or energy overhead. This paper proposes a counter architecture for online DVFS profitability estimation on superscalar out-of-order processors. The counter architecture teases apart the fraction of the execution time that is susceptible to clock frequency versus the fraction that is insusceptible to clock frequency. By doing so, the counter architecture can accurately estimate the performance and energy consumption at different V/f operating points from a single program execution. The DVFS counter architecture estimates performance, energy consumption and ED2P within 0.2%, 0.5% and 0.8% on average, respectively, over a 4× frequency range. Further, the counter architecture incurs a limited hardware cost and is an enabler for online DVFS scheduling both at the intra-core as well as at the inter-core level in a multi-core processor.
Keywords
ENERGY, Computer systems organization, performance of systems, modeling techniques, modeling of computer architecture, PROCESSOR, MICROPROCESSORS, MICROARCHITECTURE, PERFORMANCE, POWER MANAGEMENT, power management

Downloads

  • (...).pdf
    • full text
    • |
    • UGent only
    • |
    • PDF
    • |
    • 1.81 MB

Citation

Please use this url to cite or link to this publication:

Chicago
Eyerman, Stijn, and Lieven Eeckhout. 2010. “A Counter Architecture for Online DVFS Profitability Estimation.” Ieee Transactions on Computers 59 (11): 1576–1583.
APA
Eyerman, S., & Eeckhout, L. (2010). A counter architecture for online DVFS profitability estimation. IEEE TRANSACTIONS ON COMPUTERS, 59(11), 1576–1583.
Vancouver
1.
Eyerman S, Eeckhout L. A counter architecture for online DVFS profitability estimation. IEEE TRANSACTIONS ON COMPUTERS. 2010;59(11):1576–83.
MLA
Eyerman, Stijn, and Lieven Eeckhout. “A Counter Architecture for Online DVFS Profitability Estimation.” IEEE TRANSACTIONS ON COMPUTERS 59.11 (2010): 1576–1583. Print.
@article{1087974,
  abstract     = {Dynamic voltage and frequency scaling (DVFS) is a well-known and effective technique for reducing power consumption in modern microprocessors. An important concern though when applying DVFS for online energy and power optimizations is to estimate its profitability in terms of performance and energy. Current DVFS profitability estimation approaches however lack accuracy or incur runtime performance and/or energy overhead.
This paper proposes a counter architecture for online DVFS profitability estimation on superscalar out-of-order processors. The counter architecture teases apart the fraction of the execution time that is susceptible to clock frequency versus the fraction that is insusceptible to clock frequency. By doing so, the counter architecture can accurately estimate the performance and energy consumption at different V/f operating points from a single program execution. The DVFS counter architecture estimates performance, energy consumption and ED2P within 0.2\%, 0.5\% and 0.8\% on average, respectively, over a 4{\texttimes} frequency range. Further, the counter architecture incurs a limited hardware cost and is an enabler for online DVFS scheduling both at the intra-core as well as at the inter-core level in a multi-core processor.},
  author       = {Eyerman, Stijn and Eeckhout, Lieven},
  issn         = {0018-9340},
  journal      = {IEEE TRANSACTIONS ON COMPUTERS},
  keyword      = {ENERGY,Computer systems organization,performance of systems,modeling techniques,modeling of computer architecture,PROCESSOR,MICROPROCESSORS,MICROARCHITECTURE,PERFORMANCE,POWER MANAGEMENT,power management},
  language     = {eng},
  number       = {11},
  pages        = {1576--1583},
  title        = {A counter architecture for online DVFS profitability estimation},
  url          = {http://dx.doi.org/10.1109/TC.2010.65},
  volume       = {59},
  year         = {2010},
}

Altmetric
View in Altmetric
Web of Science
Times cited: