Energy Claims at Scale: Decreasing the Energy Demand of HPC Workloads at OS Level

被引:1
作者
Eibel, Christopher [1 ]
Hoenig, Timo [1 ]
Schroeder-Preikschat, Wolfgang [1 ]
机构
[1] Friedrich Alexander Universitat Erlangen Nurnberg, Erlangen, Germany
来源
2016 IEEE 30TH INTERNATIONAL PARALLEL AND DISTRIBUTED PROCESSING SYMPOSIUM WORKSHOPS (IPDPSW) | 2016年
关键词
SYSTEM;
D O I
10.1109/IPDPSW.2016.69
中图分类号
TP3 [计算技术、计算机技术];
学科分类号
0812 ;
摘要
The ever-increasing performance and power demand of HPC systems requires sophisticated approaches that improve energy-efficient job execution. Targeted goals such as the 20-MW limit for an exascale system, set by the Department of Energy (DoE), are not achievable with advances at hardware level only. Instead, it is inevitable to establish new concepts at system-software level. We propose a system software in which the operating system acquires a key position in the energy-reduction process. Our approach advocates heterogeneous hardware components-especially those that currently emerge from the embedded-system domain-to achieve both high performance and high energy efficiency.
引用
收藏
页码:1114 / 1117
页数:4
相关论文
共 19 条
  • [1] Almási G, 2003, LECT NOTES COMPUT SC, V2790, P543
  • [2] [Anonymous], 2013, NSDI
  • [3] Adaptive Configuration Selection for Power-Constrained Heterogeneous Systems
    Bailey, Peter E.
    Lowenthal, David K.
    Ravi, Vignesh
    Rountree, Barry
    Schulz, Martin
    de Supinski, Bronis R.
    [J]. 2014 43RD INTERNATIONAL CONFERENCE ON PARALLEL PROCESSING (ICPP), 2014, : 371 - 380
  • [4] DIET: A scalable toolbox to build network enabled servers on the grid
    Caron, E.
    Desprez, F.
    [J]. INTERNATIONAL JOURNAL OF HIGH PERFORMANCE COMPUTING APPLICATIONS, 2006, 20 (03) : 335 - 352
  • [5] Cirne Walfredo., 2001, IPDPS, P59
  • [6] Eibel C., 2015, P 14 WORKSH AD REFL, P19
  • [7] POWERPC-603(TM), A MICROPROCESSOR FOR PORTABLE COMPUTERS
    GARY, S
    IPPOLITO, P
    GEROSA, G
    DIETZ, C
    ENO, J
    SANCHEZ, H
    [J]. IEEE DESIGN & TEST OF COMPUTERS, 1994, 11 (04): : 14 - 23
  • [8] Energy Efficient HPC on Embedded SoCs: Optimization Techniques for Mali GPU
    Grasso, Ivan
    Radojkovic, Petar
    Rajovic, Nikola
    Gelado, Isaac
    Ramirez, Alex
    [J]. 2014 IEEE 28TH INTERNATIONAL PARALLEL AND DISTRIBUTED PROCESSING SYMPOSIUM, 2014,
  • [9] Honig T., 2014, Proceedings of the 2014 Conference on Timely Results in Operating Systems (TRIOS'14), P1
  • [10] Honig T., 2015, P 2015 BRAZ S COMP S