THERMAL MANGAMENT OF MULTICORE PROCESSORS USING POWER MULTIPLEXING

被引:0
|
作者
Gupta, Man Prakash [1 ]
Cho, Minki [2 ]
Mukhopadhyay, Saibal [2 ]
Kumar, Satish [1 ]
机构
[1] Georgia Inst Technol, Dept Mech Engn, Atlanta, GA 30332 USA
[2] Georgia Inst Technol, Dept Elect & Comp Engn, Atlanta, GA USA
关键词
Multicore; Timeslice; Multiplexing; Core hopping; Hot spots;
D O I
暂无
中图分类号
O414.1 [热力学];
学科分类号
摘要
One of the novel methods for the thermal management of multi-core processors is power multiplexing (also known as core hopping) which involves dynamical change of the locations of active cores within the chip at fixed time intervals. The power multiplexing technique helps in reducing the number of hotspots on the chip by providing a spatially uniform thermal profile which in turn lowers the maximum temperature rise on the chip. We quantify the effects of power multiplexing on the thermal profile of multi-core processor chip. Different core migration policies have been implemented in an attempt to evolve an optimally suitable policy for the multiplexing purpose. We observe that the selection of appropriate migration policy and the migration rate can efficiently reduce the spatial non-uniformity and peak temperature on the chip. The ratio of active to total cores has been varied to accommodate and analyze the effect of varying computing workload. We correlated the cooling power with the peak temperature on the chip and discussed the efficient usage of core-migration policies in the context of the power reduction.
引用
收藏
页数:7
相关论文
共 50 条
  • [21] Dynamic Power Management for Dark Silicon Multicore Processors
    Garg, Siddharth
    DARK SILICON AND FUTURE ON-CHIP SYSTEMS, 2018, 110 : 171 - 216
  • [22] Power efficiency for variation-tolerant multicore processors
    Donald, James
    Martonosi, Margaret
    ISLPED '06: PROCEEDINGS OF THE 2006 INTERNATIONAL SYMPOSIUM ON LOW POWER ELECTRONICS AND DESIGN, 2006, : 304 - 309
  • [23] Multicore processors and GPUs: the power of parallel computing in the Cloud
    Bennett, Kelly W.
    Robertson, James
    ARTIFICIAL INTELLIGENCE AND MACHINE LEARNING FOR MULTI-DOMAIN OPERATIONS APPLICATIONS II, 2020, 11413
  • [24] Supervised Learning Based Power Management for Multicore Processors
    Jung, Hwisung
    Pedram, Massoud
    IEEE TRANSACTIONS ON COMPUTER-AIDED DESIGN OF INTEGRATED CIRCUITS AND SYSTEMS, 2010, 29 (09) : 1395 - 1408
  • [25] Accurate On-Chip Temperature Sensing for Multicore Processors Using Embedded Thermal Sensors
    Li, Xin
    Li, Zhi
    Zhou, Wei
    Duan, Zhemin
    IEEE TRANSACTIONS ON VERY LARGE SCALE INTEGRATION (VLSI) SYSTEMS, 2020, 28 (11) : 2328 - 2341
  • [26] Enhanced Detection of Thermal Covert Channel Attacks in Multicore Processors
    Dhananjay, Krithika
    Pavlidis, Vasilis F.
    Coskun, Ayse K.
    Salman, Emre
    2024 25TH INTERNATIONAL SYMPOSIUM ON QUALITY ELECTRONIC DESIGN, ISQED 2024, 2024,
  • [27] Information-flow control on ARM and POWER multicore processors
    Smith, Graeme
    Coughlin, Nicholas
    Murray, Toby
    FORMAL METHODS IN SYSTEM DESIGN, 2021, 58 (1-2) : 251 - 293
  • [28] Efficient Traffic Aware Power Management for Multicore Communications Processors
    Iqbal, Muhammad Faisal
    John, Lizy K.
    PROCEEDINGS OF THE EIGHTH ACM/IEEE SYMPOSIUM ON ARCHITECTURES FOR NETWORKING AND COMMUNICATIONS SYSTEMS (ANCS'12), 2012, : 123 - 133
  • [29] Information-flow control on ARM and POWER multicore processors
    Graeme Smith
    Nicholas Coughlin
    Toby Murray
    Formal Methods in System Design, 2021, 58 : 251 - 293
  • [30] Countering power analysis attacks by exploiting characteristics of multicore processors
    Yang, Jianwei
    Dai, Fan
    Wang, Jielin
    Zeng, Jianmin
    Zhang, Zhang
    Han, Jun
    Zeng, Xiaoyang
    IEICE ELECTRONICS EXPRESS, 2018, 15 (07):