User-driven frequency scaling

Arindam Mallik; Bin Lin; Gokhan Memik; Peter A Dinda; Robert P. Dick

doi:10.1109/L-CA.2006.16

User-driven frequency scaling

Arindam Mallik^*, Bin Lin, Gokhan Memik, Peter A Dinda, Robert P. Dick

^*Corresponding author for this work

Research output: Contribution to journal › Article › peer-review

34 Scopus citations

Abstract

We propose and evaluate User-Driven Frequency Scaling (UDFS) for improved power management on processors that support Dynamic Voltage and Frequency Scaling (DVFS), e.g, those used in current laptop and desktop computers. UDFS dynamically adapts CPU frequency to the individual user and the workload through a simple user feedback mechanism, unlike currently-used DVFS methods which rely only on CPU utilization. Our UDFS algorithms dramatically reduce typical operating frequencies while maintaining performance at satisfactory levels for each user. We evaluated our techniques through user studies conducted on a Pentium M laptop running Windows applications. The UDFS scheme reduces measured system power by 22.1%, averaged across all our users and applications, compared to the Windows XP DVFS scheme.

Original language	English (US)
Journal	IEEE Computer Architecture Letters
Volume	5
Issue number	2
DOIs	https://doi.org/10.1109/L-CA.2006.16
State	Published - Feb 2006

ASJC Scopus subject areas

Hardware and Architecture

Access to Document

10.1109/L-CA.2006.16

Cite this

@article{222c4e0336c4497c81783483f42e72a5,

title = "User-driven frequency scaling",

abstract = "We propose and evaluate User-Driven Frequency Scaling (UDFS) for improved power management on processors that support Dynamic Voltage and Frequency Scaling (DVFS), e.g, those used in current laptop and desktop computers. UDFS dynamically adapts CPU frequency to the individual user and the workload through a simple user feedback mechanism, unlike currently-used DVFS methods which rely only on CPU utilization. Our UDFS algorithms dramatically reduce typical operating frequencies while maintaining performance at satisfactory levels for each user. We evaluated our techniques through user studies conducted on a Pentium M laptop running Windows applications. The UDFS scheme reduces measured system power by 22.1%, averaged across all our users and applications, compared to the Windows XP DVFS scheme.",

author = "Arindam Mallik and Bin Lin and Gokhan Memik and Dinda, {Peter A} and Dick, {Robert P.}",

note = "Funding Information: Manuscript submitted: 24 Aug. 2006. Manuscript accepted: 8 Oct. 2006. Final manuscript received: 15 Oct. 2006. This work is in part supported by Department of Energy Award DE-FG02-05ER25691 and NSF Grants IIS-0613568, CNS-0551639, CNS-0347941, CCF-0541337, IIS-0536994, CCF-0444405, ANI-0093221, ANI-0301108, and EIA-0224449.",

year = "2006",

month = feb,

doi = "10.1109/L-CA.2006.16",

language = "English (US)",

volume = "5",

journal = "IEEE Computer Architecture Letters",

issn = "1556-6056",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

number = "2",

}

TY - JOUR

T1 - User-driven frequency scaling

AU - Mallik, Arindam

AU - Lin, Bin

AU - Memik, Gokhan

AU - Dinda, Peter A

AU - Dick, Robert P.

N1 - Funding Information: Manuscript submitted: 24 Aug. 2006. Manuscript accepted: 8 Oct. 2006. Final manuscript received: 15 Oct. 2006. This work is in part supported by Department of Energy Award DE-FG02-05ER25691 and NSF Grants IIS-0613568, CNS-0551639, CNS-0347941, CCF-0541337, IIS-0536994, CCF-0444405, ANI-0093221, ANI-0301108, and EIA-0224449.

PY - 2006/2

Y1 - 2006/2

N2 - We propose and evaluate User-Driven Frequency Scaling (UDFS) for improved power management on processors that support Dynamic Voltage and Frequency Scaling (DVFS), e.g, those used in current laptop and desktop computers. UDFS dynamically adapts CPU frequency to the individual user and the workload through a simple user feedback mechanism, unlike currently-used DVFS methods which rely only on CPU utilization. Our UDFS algorithms dramatically reduce typical operating frequencies while maintaining performance at satisfactory levels for each user. We evaluated our techniques through user studies conducted on a Pentium M laptop running Windows applications. The UDFS scheme reduces measured system power by 22.1%, averaged across all our users and applications, compared to the Windows XP DVFS scheme.

AB - We propose and evaluate User-Driven Frequency Scaling (UDFS) for improved power management on processors that support Dynamic Voltage and Frequency Scaling (DVFS), e.g, those used in current laptop and desktop computers. UDFS dynamically adapts CPU frequency to the individual user and the workload through a simple user feedback mechanism, unlike currently-used DVFS methods which rely only on CPU utilization. Our UDFS algorithms dramatically reduce typical operating frequencies while maintaining performance at satisfactory levels for each user. We evaluated our techniques through user studies conducted on a Pentium M laptop running Windows applications. The UDFS scheme reduces measured system power by 22.1%, averaged across all our users and applications, compared to the Windows XP DVFS scheme.

UR - http://www.scopus.com/inward/record.url?scp=33846692739&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=33846692739&partnerID=8YFLogxK

U2 - 10.1109/L-CA.2006.16

DO - 10.1109/L-CA.2006.16

M3 - Article

AN - SCOPUS:33846692739

SN - 1556-6056

VL - 5

JO - IEEE Computer Architecture Letters

JF - IEEE Computer Architecture Letters

IS - 2

ER -

User-driven frequency scaling

Abstract

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this