Mainstream vs. Emerging HPC: Metrics, Trade-offs and Lessons Learned

被引:0
|
作者
Radulovic, Milan [1 ,2 ]
Asifuzzaman, Kazi [1 ,2 ]
Zivanovic, Darko [1 ,2 ]
Rajovic, Nikola [1 ]
de Verdiere, Guillaume Colin [3 ]
Pleiter, Dirk [4 ,5 ]
Marazakis, Manolis [6 ]
Kallimanis, Nikolaos [6 ]
Carpenter, Paul [1 ]
Radojkovic, Petar [1 ]
Ayguade, Eduard [1 ,2 ]
机构
[1] BSC, Barcelona, Spain
[2] UPC, Barcelona, Spain
[3] CEA, DAM, DIF, F-91297 Arpajon, France
[4] Univ Regensburg, Inst Theoret Phys, D-93040 Regensburg, Germany
[5] Forschungszentrum Julich, Julich Supercomp Ctr, D-52425 Julich, Germany
[6] Fdn Res & Technol Hellas FORTH, Iraklion, Greece
基金
欧盟地平线“2020”;
关键词
HPC; emerging and mainstream architectures; energy efficiency; weak vs. strong cores; memory latency and bandwidth; Byte/FLOP ratio;
D O I
10.1109/SBAC-PAD.2018.00049
中图分类号
TP3 [计算技术、计算机技术];
学科分类号
0812 ;
摘要
Various servers with different characteristics and architectures are hitting the market, and their evaluation and comparison in terms of HPC features is complex and multidimensional. In this paper, we share our experience of evaluating a diverse set of HPC systems, consisting of three mainstream and five emerging architectures. We evaluate the performance and power efficiency using prominent HPC benchmarks, High-Performance Linpack (HPL) and High Performance Conjugate Gradients (HPCG), and expand our analysis using publicly available specialized kernel benchmarks, targeting specific system components. In addition to a large body of quantitative results, we emphasize six usually overlooked aspects of the HPC platforms evaluation, and share our conclusions and lessons learned. Overall, we believe that this paper will improve the evaluation and comparison of HPC platforms, making a first step towards a more reliable and uniform methodology.
引用
收藏
页码:250 / 257
页数:8
相关论文
共 50 条
  • [21] Functional trade-offs in the limb muscles of dogs selected for running vs. fighting
    Pasi, BM
    Carrier, DR
    JOURNAL OF EVOLUTIONARY BIOLOGY, 2003, 16 (02) : 324 - 332
  • [22] Yield vs. Quality trade-offs for wheat in response to carbon dioxide and ozone
    Pleijel, Hakan
    Uddling, Johan
    GLOBAL CHANGE BIOLOGY, 2012, 18 (02) : 596 - 605
  • [23] Heuristic vs. Analytical Energy Hub Optimization: Design, Implementation, and Trade-Offs
    Grafenhorst, Simon
    Ohm, Jakob
    Poppenborg, Rafael
    Foerderer, Kevin
    Hagenmeyer, Veit
    2024 IEEE 12TH INTERNATIONAL CONFERENCE ON SMART ENERGY GRID ENGINEERING, SEGE 2024, 2024, : 145 - 150
  • [24] Efficiency vs. Immersion: Interface Design Trade-offs for an Exploratory Learning Environment
    Dragon, Toby
    Floryan, Mark
    Wilkins, Grayson
    Sparks, Thomas
    INTELLIGENT TUTORING SYSTEMS, ITS 2016, 2016, 9684 : 523 - 525
  • [25] Green Wireless Access Virtualization Implementation: Cost vs. QoS Trade-Offs
    Rahman, M. M.
    Despins, Charles
    Affes, Sofiene
    2014 3RD INTERNATIONAL CONFERENCE ON ECO-FRIENDLY COMPUTING AND COMMUNICATION SYSTEMS (ICECCS 2014), 2014, : 79 - 84
  • [26] INTERCONNECT TRADE-OFFS - CPLD VS FPGA
    AMOS, D
    ELECTRONIC ENGINEERING, 1995, 67 (819): : 81 - &
  • [27] MEMORY COST VS. CODING EFFICIENCY TRADE-OFFS FOR HEVC MOTION ESTIMATION ENGINE
    Sinangil, Mahmut E.
    Chandrakasan, Anantha P.
    Sze, Vivienne
    Zhou, Minhua
    2012 IEEE INTERNATIONAL CONFERENCE ON IMAGE PROCESSING (ICIP 2012), 2012, : 1533 - 1536
  • [28] Privacy vs. Security: Trade-Offs in the Acceptance of Smart Technologies for Aging-in-Place
    Schomakers, Eva-Maria
    Ziefle, Martina
    INTERNATIONAL JOURNAL OF HUMAN-COMPUTER INTERACTION, 2023, 39 (05) : 1043 - 1058
  • [29] Whales vs. gulls: Assessing trade-offs in wildlife and waste management in Patagonia, Argentina
    Stefanski, Stephanie E.
    Villasante, Sebastian
    ECOSYSTEM SERVICES, 2015, 16 : 294 - 305
  • [30] Predicting Efficiency/Effectiveness Trade-offs for Dense vs. Sparse Retrieval Strategy Selection
    Arabzadeh, Negar
    Yan, Xinyi
    Clarke, Charles L. A.
    PROCEEDINGS OF THE 30TH ACM INTERNATIONAL CONFERENCE ON INFORMATION & KNOWLEDGE MANAGEMENT, CIKM 2021, 2021, : 2862 - 2866