Profile-Guided Optimization of System Energy Consumption for High-Performance Operational Applications
Author:
Affiliation:

  • Article
  • | |
  • Metrics
  • |
  • Reference [13]
  • |
  • Related [20]
  • |
  • Cited by
  • | |
  • Comments
    Abstract:

    Currently many high-performance computers are used to finish operational numerical computing cyclically. The main maintenance cost originates from the large amount of electric energy, and reducing energy consumption can reduce the maintenance cost significantly. The core units for operational systems are microprocessors, and the current microprocessors prevalently support the low power technique of the dynamic voltage and frequency scaling (DVFS). DVFS reduces the energy consumption by decreasing the supply voltage and execution frequency, which generally leads to performance reduction. This paper models energy consumption of operational applications confined by time constraints, and present energy optimization techniques by DVFS in the operational systems. Differing on the way to obtain the program execution information, two energy optimization models, SEOM and CEOM are evolved. The execution information of SEOM is obtained directly from testing, and the execution information of CEOM is obtained from compiler-directed program profile. The models have been investigated in representative computer platforms, and the results show that they can save 12% the largest reduction of energy consumption.

    Reference
    [1] Feng W. Making a case for efficient supercomputing. ACM Queue, 2003,1(7):54-64. [doi: 10.1145/957717.957772]
    [2] Ge R, Feng X, Cameron K. Improvement of power-performance efficiency for high-end computing. In: Siegel HJ, ed. Proc. of the19th IEEE Int’l Parallel and Distributed Processing Symp. Washington: IEEE Computer Society, 2005. 233-234. [doi: 10.1109/IPDPS.2005.251]
    [3] Marinoni M, Elastic BG. DVFS management in processors with discrete voltage/frequency modes. IEEE Trans. on IndustrialInformatics, 2007,3(1):51-56. [doi: 10.1109/TII.2006.890494]
    [4] Banikazemi M, Poff D, Abali B. PAM: A novel performance/power aware meta-scheduler for multi-core systems. In: Teller PJ, ed.Proc. of the Int’l Conf. for High Performance Computing, Networking, Storage and Analysis (SC 2008). Washington: IEEEComputer Society, 2008. 40-54. [doi: 10.1109/SC.2008.5222643]
    [5] Zhu YF, Mueller F. DVSleak: Combining leakage reduction and voltage scaling in feedback EDF scheduling. In: Pande S, Li ZY,eds. Proc. of the ACM SIGPLAN/SIGBED 2007 Conf. on Languages, Compilers, and Tools for Embedded Systems (LCTES 2007).San Diego: ACM Press, 2007. 31-40. [doi: 10.1145/1254766.1254772]
    [6] Sampson A, Dietl W, Fortuna E, Gnanapragasam D, Ceze L, Grossman D. EnerJ: Approximate data types for safe and generallow-power computation. In: Hall MW, Padua DA, eds. Proc. of the 32nd ACM SIGPLAN Conf. on Programming Language Designand Implementation (PLDI 2011). San Diego: ACM Press, 2011. 164-174. [doi: 10.1145/1993498.1993518]
    [7] Tolentino M, Turner J, Cameron KW. Memory MISER: Improving main memory energy efficiency in servers. IEEE Trans. onComputers, 2009,58(3):336-350. [doi: 10.1109/TC.2008.177]
    [8] Shang PJ, Wang J. A novel power management for CMP systems in data-intensive environment. In: Sussman A, ed. Proc. of the25th IEEE Int’l Symp. on Parallel and Distributed Processing (IPDPS 2011). Washington: IEEE Computer Society, 2011. 92-103.[doi: 10.1109/IPDPS.2011.19]
    [9] Bianchini R, Rajamony R. Power and energy management for server systems. IEEE Computer, 2004,37(11):68-74. [doi: 10.1109/MC.2004.217]
    [10] Li D, Nikolopoulos DS, Cameron KW, de Supinski BR, Schulz M. Power-Aware MPI task aggregation prediction for high-endcomputing systems. In: Bader DA, ed. Proc. of the 24th IEEE Int’l Symp. on Parallel and Distributed Processing (IPDPS 2010).Washington: IEEE Computer Society, 2010. 1-12. [doi: 10.1109/IPDPS.2010.5470464]
    [11] Rountree B, Lowenthal DK, Funk S, Freeh VW, de Supinski BR, Schulz M. Bounding energy consumption in large-scale MPIprograms. In: Verastegui B, ed. Proc. of the ACM/IEEE Conf. on High Performance Networking and Computing (SC 2007). SanDiego: ACM Press, 2007. 1-9. [doi: 10.1145/1362622.1362688]
    [12] Ge R, Feng XZ, Feng WC, Cameron KW. CPU MISER: A performance-directed, run-time system for power-aware clusters. In: LaiTH, ed. Proc. of the Int’l Conf. on Parallel Processing (ICPP 2007). Washington: IEEE Computer Society, 2007. 1-8. [doi: 10.1109/ICPP.2007.29]
    [13] Song SW, Su CY, Ge R, Vishnu A, Cameron KW. Iso-Energy-Efficiency: An approach to power-constrained parallel computation.In: Sussman A, ed. Proc. of the 25th IEEE Int’l Symp. on Parallel and Distributed Processing (IPDPS 2011). Washington: IEEEComputer Society, 2011. 128-139. [doi: 10.1109/IPDPS.2011.22]
    Cited by
    Comments
    Comments
    分享到微博
    Submit
Get Citation

易会战,罗兆成.面向高性能业务应用的基于剖视信息的系统能耗优化.软件学报,2013,24(8):1761-1774

Copy
Share
Article Metrics
  • Abstract:3653
  • PDF: 5000
  • HTML: 0
  • Cited by: 0
History
  • Received:May 28,2012
  • Revised:November 06,2012
  • Online: July 26,2013
You are the first2033297Visitors
Copyright: Institute of Software, Chinese Academy of Sciences Beijing ICP No. 05046678-4
Address:4# South Fourth Street, Zhong Guan Cun, Beijing 100190,Postal Code:100190
Phone:010-62562563 Fax:010-62562533 Email:jos@iscas.ac.cn
Technical Support:Beijing Qinyun Technology Development Co., Ltd.

Beijing Public Network Security No. 11040202500063