Power modeling for GPU architectures using McPAT

Jieun Lim, Nagesh B. Lakshminarayana, Hyesoon Kim, William Jinho Song, Sudhakar Yalamanchili, Wonyong Sung

Research output: Contribution to journalArticle

30 Citations (Scopus)

Abstract

Graphics Processing Units (GPUs) are very popular for both graphics and general-purpose applications. Since GPUs operate many processing units and manage multiple levels of memory hierarchy, they consume a significant amount of power. Although several power models for CPUs are available, the power consumption of GPUs has not been studied much yet. In this article we develop a new power model for GPUs by utilizing McPAT, a CPU power tool. We generate initial power model data from McPAT with a detailed GPU configuration, and then adjust the models by comparing them with empirical data. We use the NVIDIA's Fermi architecture for building the power model, and our model estimates the GPU power consumption with an average error of 7.7% and 12.8% for the microbenchmarks and Merge benchmarks, respectively.

Original languageEnglish
Article number26
JournalACM Transactions on Design Automation of Electronic Systems
Volume19
Issue number3
DOIs
Publication statusPublished - 2014 Jan 1

Fingerprint

Program processors
Electric power utilization
Graphics processing unit
Data storage equipment
Processing

All Science Journal Classification (ASJC) codes

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

Cite this

Lim, Jieun ; Lakshminarayana, Nagesh B. ; Kim, Hyesoon ; Song, William Jinho ; Yalamanchili, Sudhakar ; Sung, Wonyong. / Power modeling for GPU architectures using McPAT. In: ACM Transactions on Design Automation of Electronic Systems. 2014 ; Vol. 19, No. 3.
@article{498ce8631b8a49af8f6d6c4ab1c347ee,
title = "Power modeling for GPU architectures using McPAT",
abstract = "Graphics Processing Units (GPUs) are very popular for both graphics and general-purpose applications. Since GPUs operate many processing units and manage multiple levels of memory hierarchy, they consume a significant amount of power. Although several power models for CPUs are available, the power consumption of GPUs has not been studied much yet. In this article we develop a new power model for GPUs by utilizing McPAT, a CPU power tool. We generate initial power model data from McPAT with a detailed GPU configuration, and then adjust the models by comparing them with empirical data. We use the NVIDIA's Fermi architecture for building the power model, and our model estimates the GPU power consumption with an average error of 7.7{\%} and 12.8{\%} for the microbenchmarks and Merge benchmarks, respectively.",
author = "Jieun Lim and Lakshminarayana, {Nagesh B.} and Hyesoon Kim and Song, {William Jinho} and Sudhakar Yalamanchili and Wonyong Sung",
year = "2014",
month = "1",
day = "1",
doi = "10.1145/2611758",
language = "English",
volume = "19",
journal = "ACM Transactions on Design Automation of Electronic Systems",
issn = "1084-4309",
publisher = "Association for Computing Machinery (ACM)",
number = "3",

}

Power modeling for GPU architectures using McPAT. / Lim, Jieun; Lakshminarayana, Nagesh B.; Kim, Hyesoon; Song, William Jinho; Yalamanchili, Sudhakar; Sung, Wonyong.

In: ACM Transactions on Design Automation of Electronic Systems, Vol. 19, No. 3, 26, 01.01.2014.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Power modeling for GPU architectures using McPAT

AU - Lim, Jieun

AU - Lakshminarayana, Nagesh B.

AU - Kim, Hyesoon

AU - Song, William Jinho

AU - Yalamanchili, Sudhakar

AU - Sung, Wonyong

PY - 2014/1/1

Y1 - 2014/1/1

N2 - Graphics Processing Units (GPUs) are very popular for both graphics and general-purpose applications. Since GPUs operate many processing units and manage multiple levels of memory hierarchy, they consume a significant amount of power. Although several power models for CPUs are available, the power consumption of GPUs has not been studied much yet. In this article we develop a new power model for GPUs by utilizing McPAT, a CPU power tool. We generate initial power model data from McPAT with a detailed GPU configuration, and then adjust the models by comparing them with empirical data. We use the NVIDIA's Fermi architecture for building the power model, and our model estimates the GPU power consumption with an average error of 7.7% and 12.8% for the microbenchmarks and Merge benchmarks, respectively.

AB - Graphics Processing Units (GPUs) are very popular for both graphics and general-purpose applications. Since GPUs operate many processing units and manage multiple levels of memory hierarchy, they consume a significant amount of power. Although several power models for CPUs are available, the power consumption of GPUs has not been studied much yet. In this article we develop a new power model for GPUs by utilizing McPAT, a CPU power tool. We generate initial power model data from McPAT with a detailed GPU configuration, and then adjust the models by comparing them with empirical data. We use the NVIDIA's Fermi architecture for building the power model, and our model estimates the GPU power consumption with an average error of 7.7% and 12.8% for the microbenchmarks and Merge benchmarks, respectively.

UR - http://www.scopus.com/inward/record.url?scp=84902969245&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84902969245&partnerID=8YFLogxK

U2 - 10.1145/2611758

DO - 10.1145/2611758

M3 - Article

VL - 19

JO - ACM Transactions on Design Automation of Electronic Systems

JF - ACM Transactions on Design Automation of Electronic Systems

SN - 1084-4309

IS - 3

M1 - 26

ER -