STREAM results for AMD AthlonXP 2100+

From: Wei Lin and Norbert Juffa (juffa@earthlink.net)
Date: Wed Jun 19 2002 - 01:37:34 CDT

  • Next message: h-takahara@bc.jp.nec.com: "STREAM Results on the NEC SX-6"

    Note that STREAM performance is just a few percent better than
    for a previously reported Athlon systems using DDR-266 memory.
    The bottleneck apparently is the Athlon FSB which is a 133 MHz
    double data rate channel with 2.1 GB/s of bandwidth.

    -- Norbert

    CPU: AMD AthlonXP 2100+ CPU (1733 MHz)
    motherboard: Asus A7V333
    chipset: VIA KT333 chipset
    memory: 512MB of Samsung DDR-333 ("PC2700")
    OS: Windows NT4/SP6
    compiler: Compaq Visual Fortran 6.6a, df -opt:5 -fast -arch:k7 -tune:k7

    ----------------------------------------------
     Double precision appears to have 16 digits of accuracy
     Assuming 8 bytes per DOUBLE PRECISION word
    ----------------------------------------------
     Array size = 18901234
     Offset = 0
     The total memory requirement is 432 MB
     You are running each test 10 times
     The *best* time for each test is used
     ----------------------------------------------------
     Your clock granularity/precision appears to be 10000 microseconds
     The tests below will each take a time on the order
     of 210000 microseconds
        (= 21 clock ticks)
     Increase the size of the arrays if this shows that
     you are not getting at least 20 clock ticks per test.
     ----------------------------------------------------
     WARNING -- The above is only a rough guideline.
     For best results, please be sure you know the
     precision of your system timer.
     ----------------------------------------------------
    Function Rate (MB/s) RMS time Min time Max time
    Copy: 975.5476 0.3200 0.3100 0.3300
    Scale: 975.5476 0.3110 0.3100 0.3200
    Add: 1106.4137 0.4190 0.4100 0.4200
    Triad: 1080.0705 0.4220 0.4200 0.4400
     Sum of a is = 2.179880793656184E+019
     Sum of b is = 4.359761587609160E+018
     Sum of c is = 5.813015449509505E+018



    This archive was generated by hypermail 2.1.4 : Wed Jun 26 2002 - 17:09:57 CDT