research-article

Power and thermal characterization of POWER6 system

Authors:

Victor Jiménez,

Francisco J. Cazorla,

Roberto Gioiosa,

Chen-Yong Cher,

Alper Buyuktosunoglu,

Pradip BoseAuthors Info & Claims

PACT '10: Proceedings of the 19th international conference on Parallel architectures and compilation techniques

Pages 7 - 18

https://doi.org/10.1145/1854273.1854281

Published: 11 September 2010 Publication History

Abstract

Controlling power consumption and temperature is of major concern for modern computing systems. In this work we characterize thermal behavior and power consumption of an IBM POWER6-based system. We perform the characterization at several levels: application, operating system, and hardware level, both when the system is idle, and under load. At hardware level, we report a 25% reduction in total system power consumption by using the processor low power mode. We also study the effect of the hardware thread prioritization mechanism provided by POWER6 on different workloads and how this mechanism can be used to limit power consumption. At OS level, we analyze the power reduction techniques implemented in the Linux kernel, such as the tickless kernel and the CPU idle power manager. At application level, we characterize the power consumption and the temperature of two sets of benchmarks (METbench and SPEC CPU2006) and we study the effect of workload characteristics on power consumption and core temperature.

From this characterization we derive a model based on performance counters that allows us to predict the total power consumption of the POWER6 system with an average error under 3% for CMP and 5% for SMT. To the best of our knowledge, this is the first power model of a system including CMP+SMT processors. Finally, we show that the static decision on whether to consolidate tasks into the same core/chip, as it is currently done in Linux, can be improved by dynamically considering the low-power capabilities of the underlying architecture and the characteristics of the application (up to 5X improvement in ED2P).

References

[1]

}}Power ISA Version 2.06, 2009. http://www.power.org/resources/downloads/PowerISA_V2.06_PUBLIC.pdf.

[2]

}}B. Behle et al. IBM EnergyScale for POWER6 Processor-Based Systems. IBM White Paper, 2009.

[3]

}}W. Bircher et al. Complete system power estimation: A trickle-down approach based on performance events. ISPASS, 2007.

[4]

}}C. Boneti et al. Software-Controlled Priority Characterization of POWER5 Processor. ISCA, 2008.

Digital Library

[5]

}}D. M. Brooks et al. Power-Aware Microarchitecture: Design and Modeling Challenges for Next-Generation Microprocessors. MICRO, 2000.

Digital Library

[6]

}}J. Casazza. Intel Core i7-800 Processor Series and the Intel core i5-700 Processor Series Based on Intel Microarchitecture (Nehalem), 2009.

[7]

}}J. Choi et al. Thermal-aware task scheduling at the system software level. ISLPED, 2007.

Digital Library

[8]

}}EPA. EPA Report to Congress on Server and Data Center Energy Efficiency. Technical report, U.S. Environmental Protection Agency, 2007.

[9]

}}X. Fan et al. Power provisioning for a warehouse-sized computer. ISCA, 2007.

Digital Library

[10]

}}M. S. Floyd et al. System power management support in the IBM POWER6 microprocessor. IBM J. Res. Dev., 51(6), 2007.

Digital Library

[11]

}}R. Gioiosa et al. Analysis of System Overhead on Parallel Computers. ISSPIT, 2004.

[12]

}}R. Gioiosa et al. Transparent Incremental Checkpoint at Kernel level: A Foundation for Fault Tolerance for Parallel Computers. SC, 2005.

Digital Library

[13]

}}H. Hanson et al. Power, Performance, and Thermal Management for High-Performance Systems. HPPAC, 2007.

[14]

}}H. Hanson et al. Thermal response to DVFS: Analysis with an Intel Pentium M. ISLPED, 2007.

Digital Library

[15]

}}F.E. Harrell, Jr. Regression Modeling Strategies. Springer-Verlag New York, Inc., 2006.

Digital Library

[16]

}}J. L. Henning. Spec cpu2006 benchmark descriptions. SIGARCH Computer Architecture News, 2006.

Digital Library

[17]

}}E. Kursun et al. Variation-aware thermal characterization and management of multi-core architectures. ICCD, 2008.

[18]

}}H. Q. Le et al. IBM POWER6 microarchitecture. IBM J. Res. Dev., 51(6), 2007.

Digital Library

[19]

}}Y. Li et al. Performance, Energy, and Thermal Considerations for SMT and CMP Architectures. HPCA, 2005.

Digital Library

[20]

}}D. Meisner et al. PowerNap: Eliminating server idle power. ASPLOS, 2009.

Digital Library

[21]

}}A. Naveh et al. Power and thermal management in the Intel Core Duo processor. Intel Technology Journal, 10(2), 2006.

[22]

}}V. Pallipadi. Cpuidle - Do nothing, efficiently... Linux Symposium, June 2007.

[23]

}}M. D. Powell et al. CAMP: A technique to estimate per-structure power at run-time using a few simple parameters. HPCA, 2009.

[24]

}}K. K. Pusukuri et al. A Methodology for Developing Simple and Robust Power Models Using Performance Monitoring Events. WIOSCA, 2009.

[25]

}}R. Sasanka et al. The energy efficiency of CMP vs. SMT for multimedia workloads. ICS, 2004.

Digital Library

[26]

}}S. Siddha et al. Getting maximum mileage out of tickless. Linux Symposium, June 2007.

[27]

}}B. Sinharoy et al. POWER5 system microarchitecture. IBM J. Res. Dev., 49(4/5), 2005.

Digital Library

[28]

}}V. Srinivasan et al. Energy-Aware Task and Interrupt Management in Linux. Linux Symposium, 2, August 2008.

Cited By

Shah PShenoy RSrinivasan VBose PBuyuktosunoglu A(2022)TokenSmart: Distributed, Scalable Power Management in the Many-core EraACM Transactions on Architecture and Code Optimization10.1145/355976220:1(1-26)Online publication date: 17-Nov-2022
https://dl.acm.org/doi/10.1145/3559762
Trilla DHernandez CAbella JCazorla F(2021)Worst-Case Energy Consumption: A New Challenge for Battery-Powered Critical DevicesIEEE Transactions on Sustainable Computing10.1109/TSUSC.2019.29431426:3(522-530)Online publication date: 1-Jul-2021
https://doi.org/10.1109/TSUSC.2019.2943142
Aljuhni AYusuf SEdward Chow CAkanbi OAljaedi A(2020)Proposed Design for Effectively Expand Adaptive-ticks Feature in the Linux Kernel to Full Tickless Function2020 International Conference on Computational Science and Computational Intelligence (CSCI)10.1109/CSCI51800.2020.00233(1250-1255)Online publication date: Dec-2020
https://doi.org/10.1109/CSCI51800.2020.00233
Show More Cited By

Index Terms

Power and thermal characterization of POWER6 system

Recommendations

Power Analysis of System-Level On-Chip Communication Architectures
CODES+ISSS '04: Proceedings of the international conference on Hardware/Software Codesign and System Synthesis: 2004

For complex System-on-chips (SoCs) fabricated in nanometer technologies, the system-level on-chip communication architecture is emerging as a significant source of power consumption. Managing and optimizing this important component of SoC power requires ...
Operation Tables for Scheduling in the Presence of Incomplete Bypassing
CODES+ISSS '04: Proceedings of the international conference on Hardware/Software Codesign and System Synthesis: 2004

Register bypassing is a powerful and widely used feature in modern processors to eliminate certain data hazards. Although complete bypassing is ideal for performance, bypassing has significant impact on cycle time, area, and power consumption of the ...
System level power and thermal management on embedded processors

Comments

Information & Contributors

Information

Published In

cover image ACM Conferences

PACT '10: Proceedings of the 19th international conference on Parallel architectures and compilation techniques

September 2010

596 pages

ISBN:9781450301787

DOI:10.1145/1854273

General Chair:
Valentina Salapura
IBM TJ Watson Research Center
,
Program Chairs:
Michael Gschwind
IBM Systems & Technology Group
,
Jens Knoop
Technische Universität Wien

Copyright © 2010 ACM.

Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. Copyrights for components of this work owned by others than ACM must be honored. Abstracting with credit is permitted. To copy otherwise, or republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee. Request permissions from [email protected]

Sponsors

IFIP WG 10.3: IFIP working group 10.3 on concurrent systems
IEEE CS TCPP: IEEE-CS technical committee on parallel processing
SIGARCH: ACM Special Interest Group on Computer Architecture
IEEE CS TCAA: IEEE CS technical committee on architectural acoustics

Publisher

Association for Computing Machinery

New York, NY, United States

Publication History

Published: 11 September 2010

Permissions

Request permissions for this article.

Request Permissions

Check for updates

Author Tags

Qualifiers

Research-article

Conference

PACT '10

Sponsor:

IFIP WG 10.3
IEEE CS TCPP
SIGARCH
IEEE CS TCAA

PACT '10: International Conference on Parallel Architectures and Compilation Techniques

September 11 - 15, 2010

Vienna, Austria

Acceptance Rates

Overall Acceptance Rate 121 of 471 submissions, 26%

Contributors

Other Metrics

View Article Metrics

Bibliometrics & Citations

Bibliometrics

Article Metrics

17
Total Citations
View Citations
446
Total Downloads

Downloads (Last 12 months)7
Downloads (Last 6 weeks)3

Reflects downloads up to 09 Nov 2024

Other Metrics

View Author Metrics

Citations

Cited By

Shah PShenoy RSrinivasan VBose PBuyuktosunoglu A(2022)TokenSmart: Distributed, Scalable Power Management in the Many-core EraACM Transactions on Architecture and Code Optimization10.1145/355976220:1(1-26)Online publication date: 17-Nov-2022
https://dl.acm.org/doi/10.1145/3559762
Trilla DHernandez CAbella JCazorla F(2021)Worst-Case Energy Consumption: A New Challenge for Battery-Powered Critical DevicesIEEE Transactions on Sustainable Computing10.1109/TSUSC.2019.29431426:3(522-530)Online publication date: 1-Jul-2021
https://doi.org/10.1109/TSUSC.2019.2943142
Aljuhni AYusuf SEdward Chow CAkanbi OAljaedi A(2020)Proposed Design for Effectively Expand Adaptive-ticks Feature in the Linux Kernel to Full Tickless Function2020 International Conference on Computational Science and Computational Intelligence (CSCI)10.1109/CSCI51800.2020.00233(1250-1255)Online publication date: Dec-2020
https://doi.org/10.1109/CSCI51800.2020.00233
Aljuhni AChow CAljaedi AYusuf STorres-Reyes F(2018)Towards understanding application performance and system behavior with the full dynticks feature2018 IEEE 8th Annual Computing and Communication Workshop and Conference (CCWC)10.1109/CCWC.2018.8301733(394-401)Online publication date: Jan-2018
https://doi.org/10.1109/CCWC.2018.8301733
Chatzopoulos GDragojević AGuerraoui R(2017)ESTIMAACM Transactions on Parallel Computing10.1145/31081374:2(1-28)Online publication date: 22-Aug-2017
https://dl.acm.org/doi/10.1145/3108137
Chatzopoulos GDragojević AGuerraoui R(2016)ESTIMAACM SIGPLAN Notices10.1145/3016078.285115951:8(1-11)Online publication date: 27-Feb-2016
https://dl.acm.org/doi/10.1145/3016078.2851159
Chatzopoulos GDragojević AGuerraoui RAsenjo RHarris T(2016)ESTIMAProceedings of the 21st ACM SIGPLAN Symposium on Principles and Practice of Parallel Programming10.1145/2851141.2851159(1-11)Online publication date: 27-Feb-2016
https://dl.acm.org/doi/10.1145/2851141.2851159
Dayarathna MWen YFan R(2016)Data Center Energy Consumption Modeling: A SurveyIEEE Communications Surveys & Tutorials10.1109/COMST.2015.248118318:1(732-794)Online publication date: Sep-2017
https://doi.org/10.1109/COMST.2015.2481183
Zhang ZLang MPakin SFu S(2016)TracSimParallel Computing10.1016/j.parco.2015.11.00257:C(108-124)Online publication date: 1-Sep-2016
https://dl.acm.org/doi/10.1016/j.parco.2015.11.002
Zhang ZLang MPakin SFu SCameron KHoisie AKerbyson DLowenthal DNikolopoulos DYalamanchili SMarquez A(2014)Trapped capacityProceedings of the 2nd International Workshop on Energy Efficient Supercomputing10.1109/E2SC.2014.10(41-50)Online publication date: 16-Nov-2014
https://dl.acm.org/doi/10.1109/E2SC.2014.10
Show More Cited By

View Options

Get Access

Login options

Check if you have access through your login credentials or your institution to get full access on this article.

Full Access

Get this Publication

View options

PDF

View or Download as a PDF file.

eReader

View online with eReader.

Media

Figures

Other

Tables

View Table of Contents