Hi.
I found one broken link in:
http://www.cs.virginia.edu/stream/
The link to the Linux version is:
ftp://ftp.cs.virginias.edu/pub/stream/Contrib/MasonCabot/linux/stream_l
while it should be: ^
ftp://ftp.cs.virginia.edu/pub/stream/Contrib/MasonCabot/linux/stream_l
Also, when I tried to execute it I got:
./stream_l: error in loading shared libraries: libstdc++-libc6.0-1.so.2:
cannot open shared object file: No such file or directory
This is because my system only has more recent versions:
# cd /usr/lib ; ls -l libcstdc++* | cut -c55-
libstdc++-2-libc6.1-1-2.9.0.a
libstdc++-2-libc6.1-1-2.9.0.so
libstdc++-libc6.1-1.a.2 -> libstdc++-2-libc6.1-1-2.9.0.a
libstdc++-libc6.1-1.so.2 -> libstdc++-2-libc6.1-1-2.9.0.so
You might want to make available a static linked version, if you want to
make it very easy to run.
Creating a "fake" link solved it:
/usr/lib # ln -s libstdc++-2-libc6.1-1-2.9.0.so libstdc++-libc6.0-1.so.2
BTW, my results:
% dmesg
Linux version 2.2.13
CPU: AMD AMD-K7(tm) Processor stepping 02 600 MHz
motherboard: FIC SD-11 (AMD 751 North Bridge, VIA 686A South Bridge)
memory: 100 MHz, 64 MB
% ./stream_l 600 1000
-------------------------------------------------------------
This system uses 8 bytes per DOUBLE PRECISION word.
-------------------------------------------------------------
Array size = 999936, Offset = 0
Total memory required = 24.9 MB.
Each test is run 1000 times, but only
the *best* time for each is used.
-------------------------------------------------------------
Your clock granularity/precision appears to be 1 microseconds.
(I have 7 such systems, so the following is the best line of each
computer.)
res_castor2:Function Rate (MB/s) RMS time Min time Max time
res_castor2:Copy: 327.0151 0.0502 0.0489 0.0505
res_castor3:Scale: 409.6746 0.0392 0.0391 0.0404
res_castor6:Add: 409.4586 0.0589 0.0586 0.0592
res_castor4:Triad: 417.2070 0.0578 0.0575 0.0581
res_castor2:CPU: AMD AMD-K7(tm) Processor stepping 02
res_castor3:CPU: AMD AMD-K7(tm) Processor stepping 02
res_castor4:CPU: AMD AMD Athlon(tm) Processor stepping 01
res_castor6:CPU: AMD AMD-K7(tm) Processor stepping 02
Similar systems, but with 550 MHz processors:
./stream_l 550 10
res_castor13:Function Rate (MB/s) RMS time Min time Max time
res_castor9:Copy: 348.4000 0.0469 0.0459 0.0473
res_castor13:Scale: 420.0487 0.0383 0.0381 0.0386
res_castor13:Add: 421.5415 0.0572 0.0569 0.0576
res_castor13:Triad: 413.5224 0.0583 0.0580 0.0588
res_castor9:CPU: AMD AMD-K7(tm) Processor stepping 02
res_castor13:CPU: AMD AMD-K7(tm) Processor stepping 02
(Curiously only the Triad time is better on the 600 MHz. In the other
tests the 550 is faster.)
Pentium II 266 MHz, LX chipset, 66MHz memory
./stream_l 266 20
-------------------------------------------------------------
This system uses 8 bytes per DOUBLE PRECISION word.
-------------------------------------------------------------
Array size = 999936, Offset = 0
Total memory required = 24.9 MB.
Each test is run 20 times, but only
the *best* time for each is used.
-------------------------------------------------------------
Your clock granularity/precision appears to be 1 microseconds.
Function Rate (MB/s) RMS time Min time Max time
Copy: 132.2413 0.1702 0.1210 0.2492
Scale: 156.9061 0.1022 0.1020 0.1026
Add: 166.4182 0.1446 0.1442 0.1448
Triad: 164.8150 0.1460 0.1456 0.1471
-- rps
This archive was generated by hypermail 2b29 : Tue Apr 25 2000 - 01:49:24 CDT