Thermal and Power Challenges in High Performance Computing Systems

被引:9
|
作者
Natarajan, Venkat [1 ]
Deshpande, Anand [1 ]
Solanki, Sudarshan [1 ]
Chandrasekhar, Arun [1 ]
机构
[1] Intel Technol India Pvt Ltd, Bangalore 560103, Karnataka, India
关键词
LOCAL HEAT-TRANSFER; PRESSURE-DROP; MASS-TRANSFER; SQUARE PRISM; ON-PACKAGE; SURFACE; ARRAYS;
D O I
10.1143/JJAP.48.05EA01
中图分类号
O59 [应用物理学];
学科分类号
摘要
This paper provides an overview of the thermal and power challenges in emerging high performance computing platforms. The advent of new sophisticated applications in highly diverse areas such as health, education, finance, entertainment, etc. is driving the platform and device requirements for future systems. The key ingredients of future platforms are vertically integrated (3D) die-stacked devices which provide the required performance characteristics with the associated form factor advantages. Two of the major challenges to the design of through silicon via (TSV) based 3D stacked technologies are (i) effective thermal management and (ii) efficient power delivery mechanisms. Some of the key challenges that are articulated in this paper include hot-spot superposition and intensification in a 3D stack, design/optimization of thermal through silicon vias (TTSVs), non-uniform power loading of multi-die stacks, efficient on-chip power delivery, minimization of electrical hotspots etc. (C) 2009 The Japan Society of Applied Physics
引用
收藏
页数:6
相关论文
共 50 条
  • [41] High-performance computing and machine learning applied in thermal systems analysis Preface
    Safdari Shadloo, Mostafa
    Rahmat, Amin
    Li, Larry K. B.
    Mahian, Omid
    Alagumalai, Avinash
    JOURNAL OF THERMAL ANALYSIS AND CALORIMETRY, 2021, 145 (04) : 1733 - 1737
  • [42] High performance computing in power system applications
    Falcao, DM
    VECTOR AND PARALLEL PROCESSING - VECPAR'96, 1997, 1215 : 1 - 23
  • [43] Utilizing the power of high-performance computing
    Liu, WH
    Prasanna, VK
    IEEE SIGNAL PROCESSING MAGAZINE, 1998, 15 (05) : 85 - 100
  • [44] Trends in high performance computing with low power
    Nakamura, Tadao
    INFORMATION-AN INTERNATIONAL INTERDISCIPLINARY JOURNAL, 2007, 10 (05): : 587 - 596
  • [45] Survey of Methodologies, Approaches, and Challenges in Parallel Programming Using High-Performance Computing Systems
    Czarnul, Pawel
    Proficz, Jerzy
    Drypczewski, Krzysztof
    SCIENTIFIC PROGRAMMING, 2020, 2020
  • [46] Applications and Trends of High Performance Computing for Electric Power Systems: Focusing on Smart Grid
    Green, Robert C., II
    Wang, Lingfeng
    Alam, Mansoor
    IEEE TRANSACTIONS ON SMART GRID, 2013, 4 (02) : 922 - 931
  • [47] Vertical Power Delivery for High Performance Computing Systems with Buck-Derived Regulators
    Krishnakumar, Sriharini
    Choi, Mingeun
    Khorasani, Ramin Rahimzadeh
    Sharma, Rohit
    Swaminathan, Madhavan
    Kumar, Satish
    Partin-Vaisband, Inna
    PROCEEDINGS OF THE IEEE 74TH ELECTRONIC COMPONENTS AND TECHNOLOGY CONFERENCE, ECTC 2024, 2024, : 2136 - 2142
  • [48] Ultralow-power silicon photonic interconnect for high-performance computing systems
    Li, Guoliang
    Zheng, Xuezhe
    Lexau, Jon
    Luo, Ying
    Thacker, Hiren
    Pinguet, Thierry
    Dong, Po
    Feng, Dazeng
    Liao, Shirong
    Shafiiha, Roshanak
    Asghari, Mehdi
    Yao, Jin
    Shi, Jing
    Shubin, Ivan N.
    Patil, Dinesh
    Liu, Frankie
    Raj, Kannan
    Ho, Ron
    Cunningham, John E.
    Krishnamoorthy, Ashok V.
    OPTOELECTRONIC INTERCONNECTS AND COMPONENT INTEGRATION IX, 2010, 7607
  • [49] Signal and Power Integrity Design Methodology for High-Performance Flight Computing Systems
    Franconi, Nicholas G.
    George, Alan D.
    Geist, Alessandro D.
    Albaijes, Dennis
    2021 IEEE SPACE COMPUTING CONFERENCE (SCC), 2021, : 27 - 38
  • [50] Performance analysis challenges and framework for high-performance reconfigurable computing
    Koehler, Seth
    Curreri, John
    George, Alan D.
    PARALLEL COMPUTING, 2008, 34 (4-5) : 217 - 230