More Web Proxy on the site http://driver.im/

research-article

Identifying the optimal energy-efficient operating points of parallel workloads

Authors:

Sherief RedaAuthors Info & Claims

ICCAD '11: Proceedings of the International Conference on Computer-Aided Design

Pages 608 - 615

Published: 07 November 2011 Publication History

Abstract

As the number of cores per processor grows, there is a strong incentive to develop parallel workloads to take advantage of the hardware parallelism. In comparison to single-threaded applications, parallel workloads are more complex to characterize due to thread interactions and resource stalls. This paper presents an accurate and scalable method for determining the optimal system operating points (i.e., number of threads and DVFS settings) at runtime for parallel workloads under a set of objective functions and constraints that optimize for energy efficiency in multi-core processors. Using an extensive training data set gathered for a wide range of parallel workloads on a commercial multi-core system, we construct multinomial logistic regression (MLR) models that estimate the optimal system settings as a function of workload characteristics. We use L₁-regularization to automatically determine the relevant workload metrics for energy optimization. At runtime, our technique determines the optimal number of threads and the DVFS setting with negligible overhead. Our experiments demonstrate that our method outperforms prior techniques with up to 51% improved decision accuracy. This translates to up to 10.6% average improvement in energy-performance operation, with a maximum improvement of 30.9%. Our technique also demonstrates superior scalability as the number of potential system operating points increases.

References

[1]

A. Azevedo, I. Issenin, R. Cornea, R. Gupta, N. Dutt, A. Veidenbaum, and A. Nicolau. Profile-based dynamic voltage scheduling using program checkpoints. In Proceedings of Design, Automation and Test in Europe Conference, 2002.

Digital Library

[2]

C. Bienia and K. Li. Parsec 2.0: A new benchmark suite for chip-multiprocessors. In Proceedings of the 5th Annual Workshop on Modeling, Benchmarking and Simulation, June 2009.

[3]

K. Choi, R. Soma, and M. Pedram. Dynamic voltage and frequency scaling based on workload decomposition. In Proceedings of international symposium on Low power electronics and design, 2004.

Digital Library

[4]

M. Curtis-Maury, F. Blagojevic, C. D. Antonopoulos, and D. S. Nikolopoulos. Prediction-based power-performance adaptation of multithreaded scientific codes. IEEE Trans. Parallel Distrib. Syst., 19:1396--1410, October 2008.

Digital Library

[5]

G. Dhiman and T. S. Rosing. Dynamic voltage frequency scaling for multi-tasking systems using online learning. In Proceedings of International Symposium on Low PowerElectronics and Design, 2007.

Digital Library

[6]

M. Etinski, J. Corbalan, J. Labarta, and M. Valero. Optimizing job performance under a given power constraint in hpc centers. In Proceedings of the International Conference on Green Computing, 2010.

Digital Library

[7]

S. Herbert and D. Marculescu. Analysis of dynamic voltage/frequency scaling in chip-multiprocessors. In Proceedings of International Symposium on Low PowerElectronics and Design, 2007.

Digital Library

[8]

C. Isci, A. Buyuktosunoglu, C.-Y. Cher, P. Bose, and M. Martonosi. An analysis of efficient multi-core global power management policies: Maximizing performance for a given power budget. In Proceedings of the 39th International Symposium on Microarchitecture, 2006.

Digital Library

[9]

C. Isci, G. Contreras, and M. Martonosi. Live, runtime phase monitoring and prediction on real systems with application to dynamic power management. In Proceedings of the 39th International Symposium on Microarchitecture, 2006.

Digital Library

[10]

W. Kim, M. S. Gupta, G. yeon Wei, and D. Brooks. System level analysis of fast, per-core dvfs using on-chip switching regulators. In International Symposium on High-Performance Computer Architecture, 2008.

[11]

J. G. Koomey. Worldwide electricity used in data centers. Environmental Research Letters, 3(3):034008, 2008.

[12]

Li and J. Martinez. Dynamic power-performance adaptation of parallel computation on chip multiprocessors. In International Symposium on High-Performance Computer Architecture, 2006.

[13]

G. Magklis, M. L. Scott, G. Semeraro, D. H. Albonesi, and S. Dropsho. Profile-based dynamic voltage and frequency scaling for a multiple clock domain microprocessor. In International Symposium on Computer Architecture, 2003.

Digital Library

[14]

D. Meisner, B. T. Gold, and T. F. Wenisch. Powernap: eliminating server idle power. In Proceeding of the 14th international conference on Architectural support for programming languages and operating systems, ASPLOS '09, 2009.

Digital Library

[15]

M. Moeng and R. G. Melhem. Applying statistical machine learning to multicore voltage & frequency scaling. In Conf. Computing Frontiers, 2010.

Digital Library

[16]

G. J. Narlikar and G. E. Blelloch. Pthreads for dynamic and irregular parallelism. In In Proc. of Supercomputing Õ98, pages 4--1. IEEE, 1998.

Digital Library

[17]

D. Shin, J. Kim, and S. Lee. Low-energy intra-task voltage scheduling using static timing analysis. In Proceedings of the 38th Conference on Design Automation, 2001.

Digital Library

[18]

K. Singh, M. Bhadauria, and S. A. McKee. Real time power estimation and thread scheduling via performance counters. SIGARCH Computer Architecture News, 2009.

Digital Library

[19]

R. Teodorescu and J. Torrellas. Variation-aware application scheduling and power management for chip multiprocessors. In International Symposium on High-Performance Computer Architecture, 2008.

Digital Library

[20]

H. Yu, B. Veeravalli, and Y. Ha. Dynamic scheduling of imprecise-computation tasks in maximizing qos under energy constraints for embedded systems. In Proceedings of the 2008 Asia and South Pacific Design Automation Conference, 2008.

Digital Library

[21]

Y Zhu and F. Mueller. Feedback edf scheduling of realtime tasks exploiting dynamic voltage scaling. Real-Time Systems Journal, 2005.

Digital Library

Cited By

Kumaraswamy MGerndt M(2020)Exploiting Dynamism in HPC Applications to Optimize Energy-EfficiencyWorkshop Proceedings of the 49th International Conference on Parallel Processing10.1145/3409390.3409399(1-10)Online publication date: 17-Aug-2020
https://dl.acm.org/doi/10.1145/3409390.3409399
Salinas-Hilburg JZapater MRisco-Martín JMoya JAyala JFanucci LTeich J(2016)Unsupervised power modeling of co-allocated workloads for energy efficiency in data centersProceedings of the 2016 Conference on Design, Automation & Test in Europe10.5555/2971808.2972121(1345-1350)Online publication date: 14-Mar-2016
https://dl.acm.org/doi/10.5555/2971808.2972121
De Sensi DTorquati MDanelutto M(2016)A Reconfiguration Algorithm for Power-Aware Parallel ApplicationsACM Transactions on Architecture and Code Optimization10.1145/300405413:4(1-25)Online publication date: 2-Dec-2016
https://dl.acm.org/doi/10.1145/3004054
Show More Cited By

Identifying the optimal energy-efficient operating points of parallel workloads
1. General and reference
  1. Cross-computing tools and techniques

Recommendations

Identifying the optimal energy-efficient operating points of parallel workloads
ICCAD '11: Proceedings of the 2011 IEEE/ACM International Conference on Computer-Aided Design

As the number of cores per processor grows, there is a strong incentive to develop parallel workloads to take advantage of the hardware parallelism. In comparison to single-threaded applications, parallel workloads are more complex to characterize due ...
Energy-efficient scheduling for memory-intensive GPGPU workloads
DATE '14: Proceedings of the conference on Design, Automation & Test in Europe

High performance for a GPGPU workload is obtained by maximizing parallelism and fully utilizing the available resources. However, this is not necessarily energy efficient, especially for memory-intensive GPGPU workloads. In this work, we propose ...
Energy efficient scheduling of MapReduce workloads on heterogeneous clusters
GCM '11: Green Computing Middleware on Proceedings of the 2nd International Workshop

Energy efficiency has become the center of attention in emerging data center infrastructures as increasing energy costs continue to outgrow all other operating expenditures. In this work we investigate energy aware scheduling heuristics to increase the ...

Comments

Please enable JavaScript to view thecomments powered by Disqus.

Information & Contributors

Information

Published In

cover image ACM Conferences

ICCAD '11: Proceedings of the International Conference on Computer-Aided Design

November 2011

844 pages

ISBN:9781457713989

General Chair:
Joel Phillips
Cadence Research Labs, Berkeley, CA
,
Program Chairs:
Alan J. Hu
Univ. of British Columbia, Vancouver, BC, Canada
,
Helmut Graeb
Technische Universitaet, Munich, Germany

Sponsors

SIGDA: ACM Special Interest Group on Design Automation
IEEE-CEDA

Publisher

IEEE Press

Publication History

Published: 07 November 2011

Check for updates

Qualifiers

Research-article

Conference

ICCAD '11

Sponsor:

SIGDA

ICCAD '11: The International Conference on Computer-Aided Design

November 7 - 10, 2011

California, San Jose

Acceptance Rates

Overall Acceptance Rate 457 of 1,762 submissions, 26%

Contributors

Other Metrics

View Article Metrics

Bibliometrics & Citations

Bibliometrics

Article Metrics

8
Total Citations
View Citations
79
Total Downloads

Downloads (Last 12 months)2
Downloads (Last 6 weeks)2

Reflects downloads up to 01 Jan 2025

Other Metrics

View Author Metrics

Citations

Cited By

Kumaraswamy MGerndt M(2020)Exploiting Dynamism in HPC Applications to Optimize Energy-EfficiencyWorkshop Proceedings of the 49th International Conference on Parallel Processing10.1145/3409390.3409399(1-10)Online publication date: 17-Aug-2020
https://dl.acm.org/doi/10.1145/3409390.3409399
Salinas-Hilburg JZapater MRisco-Martín JMoya JAyala JFanucci LTeich J(2016)Unsupervised power modeling of co-allocated workloads for energy efficiency in data centersProceedings of the 2016 Conference on Design, Automation & Test in Europe10.5555/2971808.2972121(1345-1350)Online publication date: 14-Mar-2016
https://dl.acm.org/doi/10.5555/2971808.2972121
De Sensi DTorquati MDanelutto M(2016)A Reconfiguration Algorithm for Power-Aware Parallel ApplicationsACM Transactions on Architecture and Code Optimization10.1145/300405413:4(1-25)Online publication date: 2-Dec-2016
https://dl.acm.org/doi/10.1145/3004054
Barranco RTeller PGerndt MGaither K(2016)Minimization of Xeon Phi Core Use with Negligible Execution Time ImpactProceedings of the XSEDE16 Conference on Diversity, Big Data, and Science at Scale10.1145/2949550.2949581(1-8)Online publication date: 17-Jul-2016
https://dl.acm.org/doi/10.1145/2949550.2949581
Das AAl-Hashimi BMerrett G(2016)Adaptive and Hierarchical Runtime Manager for Energy-Aware Thermal Management of Embedded SystemsACM Transactions on Embedded Computing Systems10.1145/283412015:2(1-25)Online publication date: 29-Jan-2016
https://dl.acm.org/doi/10.1145/2834120
Das AKumar AVeeravalli BShafik RMerrett GAl-Hashimi BNebel WAtienza D(2015)Workload uncertainty characterization and adaptive frequency scaling for energy minimization of embedded systemsProceedings of the 2015 Design, Automation & Test in Europe Conference & Exhibition10.5555/2755753.2755764(43-48)Online publication date: 9-Mar-2015
https://dl.acm.org/doi/10.5555/2755753.2755764
Zhan XReda S(2013)Techniques for energy-efficient power budgeting in data centersProceedings of the 50th Annual Design Automation Conference10.1145/2463209.2488951(1-7)Online publication date: 29-May-2013
https://dl.acm.org/doi/10.1145/2463209.2488951
Meng JKawakami KCoskun AGroeneveld PSciuto DHassoun S(2012)Optimizing energy efficiency of 3-D multicore systems with stacked DRAM under power and thermal constraintsProceedings of the 49th Annual Design Automation Conference10.1145/2228360.2228477(648-655)Online publication date: 3-Jun-2012
https://dl.acm.org/doi/10.1145/2228360.2228477

View Options

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