A scalable framework for online power modelling of high-performance computing nodes in production

被引:3
|
作者
Pittino, Federico [1 ]
Beneventi, Francesco [1 ]
Bartolini, Andrea [1 ]
Benini, Luca [1 ,2 ]
机构
[1] Univ Bologna, Dept Elect Elect & Informat Engn DEI, Bologna, Italy
[2] Swiss Fed Inst Technol, Integrated Syst Lab, Zurich, Switzerland
来源
PROCEEDINGS 2018 INTERNATIONAL CONFERENCE ON HIGH PERFORMANCE COMPUTING & SIMULATION (HPCS) | 2018年
关键词
power model; HPC cluster in production; machine learning; scalable framework;
D O I
10.1109/HPCS.2018.00058
中图分类号
TP301 [理论、方法];
学科分类号
081202 ;
摘要
Power and thermal design and management are critical components of high performance computing (HPC) systems, due to their cutting-edge position in terms of high power density and large total power consumption. Many HPC power management strategies rely on the availability of accurate compact power models, capable of predicting power consumption and tracking its sensitivity to workload parameters and operating points. In this paper we describe a methodology and a framework for training power models derived with two of the best-in-class procedures directly on the online in production nodes and without requiring dedicated training instances. The compact power models are obtained using an online regression-based approach which can track non-stationary workloads and hardware variability. Our experiments on a real-life HPC system demonstrate that the models achieve very high accuracy over all operating modes. We also demonstrate the scalability of our approach and the small amount of resources needed for the online modeling, for both the training and inference phases.
引用
收藏
页码:300 / 307
页数:8
相关论文
共 50 条
  • [41] A Vertical Power Delivery Architecture for High-Performance Computing
    Kong, Cai
    Huan, Weiwei
    Wang, Jian
    Li, Dali
    Sun, Hui
    Zhang, Xuehong
    Ye, Fenghua
    Lin, Kaizhi
    IEEE TRANSACTIONS ON POWER ELECTRONICS, 2025, 40 (05) : 6663 - 6674
  • [42] LOW POWER HIGH-PERFORMANCE COMPUTING ON THE BEAGLEBOARD PLATFORM
    Principi, Emanuele
    Colagiacomo, Vito
    Squartini, Stefano
    Piazza, Francesco
    2012 5TH EUROPEAN DSP EDUCATION AND RESEARCH CONFERENCE (EDERC), 2012, : 35 - 39
  • [43] High-Performance Genomic Analysis Framework with In-Memory Computing
    Li, Xueqi
    Tan, Guangming
    Wang, Bingchen
    Sun, Ninghui
    ACM SIGPLAN NOTICES, 2018, 53 (01) : 317 - +
  • [44] GWmodelS: A High-Performance Computing Framework for Geographically Weighted Models
    Lu, Binbin
    Dong, Guanpeng
    SPATIAL DATA AND INTELLIGENCE, SPATIALDI 2022, 2022, 13614 : 154 - 161
  • [45] PALM:: a computational framework for assembling high-performance computing applications
    Buis, S
    Piacentini, A
    Déclat, D
    CONCURRENCY AND COMPUTATION-PRACTICE & EXPERIENCE, 2006, 18 (02): : 231 - 245
  • [46] Payload fragmentation framework for high-performance computing in cloud environment
    V. Vivek
    R. Srinivasan
    R. Elijah Blessing
    R. Dhanasekaran
    The Journal of Supercomputing, 2019, 75 : 2789 - 2804
  • [47] Payload fragmentation framework for high-performance computing in cloud environment
    Vivek, V.
    Srinivasan, R.
    Blessing, R. Elijah
    Dhanasekaran, R.
    JOURNAL OF SUPERCOMPUTING, 2019, 75 (05): : 2789 - 2804
  • [48] CGLX: A Scalable, High-Performance Visualization Framework for Networked Display Environments
    Doerr, Kai-Uwe
    Kuester, Falko
    IEEE TRANSACTIONS ON VISUALIZATION AND COMPUTER GRAPHICS, 2011, 17 (03) : 320 - 332
  • [49] High-Performance Computing
    Bungartz, Hans-Joachim
    IT-INFORMATION TECHNOLOGY, 2013, 55 (03): : 83 - 85