View page
Talk page

Mark 6 performance

This page can't be edited. | Page History
    Table of contents
    You are currently comparing two old versions - only when you are comparing against the latest version can you revert. Return to version archive.

    Combined revision comparison

    Comparing version 22:59, 14 Sep 2023 by rottmann with version 11:31, 18 Sep 2023 by jwagner.

    ...

    Testing also carried out at Bonn with e22c20 b4 scans 1026, 1033, 1070. The number of nodes was kept constant. Only v2d SETUP numBufferedFFTs and subintNS were altered.

    1. run: 1.6 ms subint, 20 buffered FFTs

    e22c20-1-b4_1019 : 748.983 sec, 12.6x slowdown, MpiDone
    e22c20-1-b4_1026 : 165.502 sec, 2.6x slowdown, MpiDone
    e22c20-1-b4_1033 : 147.769 sec, 2.5x slowdown, MpiDone
    e22c20-1-b4_1070 : 2267.63 sec, 7.6x slowdown, MpiDone

    ...

    Version from 22:59, 14 Sep 2023

    This revision modified by rottmann (Ban)

    ...

    Version as of 11:31, 18 Sep 2023

    This revision modified by jwagner (Ban)

    ...

    Testing also carried out at Bonn with e22c20 b4 scans 1026, 1033, 1070. The number of nodes was kept constant. Only v2d SETUP numBufferedFFTs and subintNS were altered.

    1. run: 1.6 ms subint, 20 buffered FFTs

    e22c20-1-b4_1019 : 748.983 sec, 12.6x slowdown, MpiDone
    e22c20-1-b4_1026 : 165.502 sec, 2.6x slowdown, MpiDone
    e22c20-1-b4_1033 : 147.769 sec, 2.5x slowdown, MpiDone
    e22c20-1-b4_1070 : 2267.63 sec, 7.6x slowdown, MpiDone

    ...