STREAM on Origin 128pe

From: Yoshio Ashizawa (
Date: Fri Jun 27 1997 - 01:28:33 CDT

Hi John,
128pe Origin system is up today in Eagan and I have
a chance to run the STREAM for the HGC benchmark.
As you instructed me, I modify the size to 80Mwords.
We are not allowed to modify the source or add the
parallel directives under this benchmark.
I parallelized the loop with -pfa. If I specify
larger size, the program will dump a core file.
(The total memory usage is over 2 GB, then the program
will fail. I do not know why.)

Compile Flags:
F77 = f77
FFLAGS = -64 -mips4 -c -O3 -pfa -i8 -listing
CC = cc
CFLAGS = -64 -mips4 -c
LFLAGS = -64 -mips4 -mp
I think I will have better numbers if I specify
more options. But only the COPY is the requirement
in this benchmark.
Thank you very much and just for your information.

+ stream
 Double precision appears to have 16 digits of accuracy
 Assuming 8 bytes per DOUBLE PRECISION word
 Array size = 80000000
 Offset = 0
 The total memory requirement is 1831 MB
 You are running each test 10 times
 The *best* time for each test is used
 Your clock granularity/precision appears to be 7 microseconds
 The tests below will each take a time on the order
 of 54419 microseconds
    (= 7774 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: 20535.1874 0.0886 0.0623 0.1336
Scale: 19554.5058 0.0925 0.0655 0.1398
Add: 23020.1977 0.0950 0.0834 0.1348
Triad: 23174.9509 0.1020 0.0828 0.1630
 Sum of a is = 9.2264047738619265E+19
 Sum of b is = 1.8452812500180128E+19
 Sum of c is = 2.4603745079650111E+19


------------------------------------------------------------------ Yoshio Ashizawa Scalable Systems Technology Center, Nihon SiliconGraphics Cray K.K. E-mail: PHONE : +81-3-3239-5101 FAX: +81-3-3239-1323

This archive was generated by hypermail 2b29 : Tue Apr 18 2000 - 05:23:07 CDT