November 2019 HPCG Results

November 2019 HPCG Results

HPCG Performance List to be announced at a SC 2019

The HPCG results continue to be integrated into the TOP500 list. As before, the 12th edition of the HPCG list of results contains all submissions submitted to the HPCG site, some of which might have not made the TOP500 list.

The 12th edition of the bi-annual High Performance Conjugate Gradients (HPCG) Benchmark list will be announced on November 21, 2019 at the SC19 conference in Denver, CO, USA. The session entitled HPCG Benchmark Update will take place on Thursday, November 21, 2019 in Rooms 704-706 at 12:15pm in the Colorado Convention Center.

This is the twelfth list produced for the benchmark that was designed to complement the traditional High Performance LINPACK (HPL) benchmark used as the official metric for ranking the TOP500 systems. The first HPCG list was announced at ISC 2014 over five years ago, containing only 15 entries. Since then, HPCG list was released bi-annually and the number of entries has steadily increased as the new results became available. The current list contains over 150 entries as HPCG continues to gain traction in the HPC community.

It is now customary that the new list contains entries from many of the TOP500's highest ranked systems, mainly the top-50. The current list is no exception. However, the HPCG ranking continues to provide a significantly different ordering of the machines when compared with the TOP500 ranking.

Summary highlights from the 10th HPCG list include:

  • Decommissioning of K computer changed the dynamic of top 5 systems.
  • Top 3 systems come from US now and were commissioned by DOE.
  • France now has a system in top 10.
  • Additional ARM systems are now part of the list.
  • A new type of system from NEC was reported in this edition and it's based SX-Aurora. Two different configurations were reported to show the scaling capability.
  • The total number of systems is 160 despite retirement of some of the machines from the previous list.

Tentative BOF schedule:

  1. Overview of the latest results
  2. Optimization updates
    1. AMD: Eelu (Lulu) Lu, Ben Sander: HPCG Benchmark
    2. ARM: Roxana Rusitoru, John Linford, Olly Perks, Dani Ruiz: HPCG on ARM, SC19 update
    3. Intel: Spencer Patty, Omid Khanmohamadi
  3. Discussion of future strategies
    1. Execution models for discrete devices
    2. Green HPCG
    3. HPCG Benchmark Advisory Group

Jun 09 2022 Contact: Admin Login