Intresting bits about IBM F50's

[gm7uac]

Just some bits of intrest about my F50

Gentoo on a F50, why bother? because I have one my machine was probably one of the last built by IBM (1999) these machine cost 20K when new the specs not to bad for it's age.

quad 332 Mhz 604r CPU's
2 GB RAM
2 * 64 bit PCI slots
5 * 32 bit PCI slots
( 2 ISA slots )
6 bay SCSI back-plane (can be expanded to 18 )
20 GB Exabyte tape drive.


SMP machine: output from proc/cpuinfo

F50 proc # cat cpuinfo
processor : 0
cpu : 604r
clock : 332MHz
revision : 49.2 (pvr 0009 3102)
bogomips : 330.75

processor : 1
cpu : 604r
clock : 332MHz
revision : 49.2 (pvr 0009 3102)
bogomips : 330.75

processor : 2
cpu : 604r
clock : 332MHz
revision : 49.2 (pvr 0009 3102)
bogomips : 330.75

processor : 3
cpu : 604r
clock : 332MHz
revision : 49.2 (pvr 0009 3102)
bogomips : 330.75

total bogomips : 1323.00
machine : CHRP IBM,7025-F50

nbench ( I know its old, but so is the machine ) the second nbench score was run while compiling KDE !
time nbench

BYTEmark* Native Mode Benchmark ver. 2 (10/95)
Index-split by Andrew D. Balsa (11/97)
Linux/Unix* port by Uwe F. Mayer (12/96,11/97)

TEST : Iterations/sec. : Old Index : New Index
: : Pentium 90* : AMD K6/233*
--------------------:------------------:-------------:------------
NUMERIC SORT : 207.52 : 5.32 : 1.75
STRING SORT : 13.219 : 5.91 : 0.91
BITFIELD : 3.7273e+07 : 6.39 : 1.34
FP EMULATION : 13.8 : 6.62 : 1.53
FOURIER : 2368.1 : 2.69 : 1.51
ASSIGNMENT : 2.4981 : 9.51 : 2.47
IDEA : 719.43 : 11.00 : 3.27
HUFFMAN : 262.66 : 7.28 : 2.33
NEURAL NET : 4.1193 : 6.62 : 2.78
LU DECOMPOSITION : 100.24 : 5.19 : 3.75
==========================ORIGINAL BYTEMARK RESULTS==========================
INTEGER INDEX : 7.211
FLOATING-POINT INDEX: 4.523
Baseline (MSDOS*) : Pentium* 90, 256 KB L2-cache, Watcom* compiler 10.0
==============================LINUX DATA BELOW===============================
CPU : 4 CPU
L2 Cache :
OS : Linux 2.6.13-gentoo-r3RS6000F50
C compiler : 3.4.4
libc :
MEMORY INDEX : 1.444
INTEGER INDEX : 2.122
FLOATING-POINT INDEX: 2.509
Baseline (LINUX) : AMD K6/233*, 512 KB L2-cache, gcc 2.7.2.3, libc-5.4.38
* Trademarks are property of their respective holder.

real 4m20.638s
user 4m20.468s
sys 0m0.097s
ed@F50 ~ $ time nbench

BYTEmark* Native Mode Benchmark ver. 2 (10/95)
Index-split by Andrew D. Balsa (11/97)
Linux/Unix* port by Uwe F. Mayer (12/96,11/97)

TEST : Iterations/sec. : Old Index : New Index
: : Pentium 90* : AMD K6/233*
--------------------:------------------:-------------:------------
NUMERIC SORT : 203.36 : 5.22 : 1.71
STRING SORT : 13.182 : 5.89 : 0.91
BITFIELD : 3.7124e+07 : 6.37 : 1.33
FP EMULATION : 13.794 : 6.62 : 1.53
FOURIER : 2365.3 : 2.69 : 1.51
ASSIGNMENT : 2.4885 : 9.47 : 2.46
IDEA : 718.56 : 10.99 : 3.26
HUFFMAN : 262.07 : 7.27 : 2.32
NEURAL NET : 4.0431 : 6.49 : 2.73
LU DECOMPOSITION : 98.16 : 5.09 : 3.67
==========================ORIGINAL BYTEMARK RESULTS==========================
INTEGER INDEX : 7.175
FLOATING-POINT INDEX: 4.462
Baseline (MSDOS*) : Pentium* 90, 256 KB L2-cache, Watcom* compiler 10.0
==============================LINUX DATA BELOW===============================
CPU : 4 CPU
L2 Cache :
OS : Linux 2.6.13-gentoo-r3RS6000F50
C compiler : 3.4.4
libc :
MEMORY INDEX : 1.439
INTEGER INDEX : 2.110
FLOATING-POINT INDEX: 2.475
Baseline (LINUX) : AMD K6/233*, 512 KB L2-cache, gcc 2.7.2.3, libc-5.4.38
* Trademarks are property of their respective holder.

real 4m21.200s
user 4m21.139s
sys 0m0.060s

The nbench results are for just 1 CPU, this app does not use mulit-threading


Just tried the boinc_client benchmark - results below

2005-12-19 20:53:06 [---] Running CPU benchmarks
2005-12-19 20:54:03 [---] Benchmark results:
2005-12-19 20:54:03 [---] Number of CPUs: 4
2005-12-19 20:54:03 [---] 168 double precision MIPS (Whetstone) per CPU
2005-12-19 20:54:03 [---] 481 integer MIPS (Dhrystone) per CPU
2005-12-19 20:54:03 [---] Finished CPU benchmarks
_________________
If all else fails RTFM.