From: Klaus.Groening@t-online.de (Klaus Groening)  
Subject: results from ubench 0.32  
Date: Sun, 18 Aug 2002 01:18:31 +0200  
   
Hello,

I use your fine ubench-0,32 -programm on my machines, here are two  results:

FreeBSD 4.6-STABLE
P IV (Northwood) 2,2 GHZ
512 MB SDRAM PC133

ubench -s =
CPU:  77,338 ; MEM:68,940;  AVG:73,179
ubench
CPU: 74,782 ; MEM:68,988;  AVG: 71,885

And the LAP  has:
FreeBSD 4.6STABLE
PIII 1,1 GHZ, 256 RAM PC100

ubench =
CPU: 62,076 ; MEM: 24,102; AVG 43,089

After the tests the first one is in CPU 20,4% better than the LAP and in
memory 186%. Okay.
I use another test under win-os about Dhrystones und Whetstones and Mem:
the first computer  has

1,975,400 KDhrystones /s and
7,236 KWhetstones /s
Mem: 319, 950 KByte /s

the Lap has
1,441,638 KDhrystones /s
7,288 KWhetstones /s
Mem: 192,021 KByte /s

No, differenz at the floating-point-test (Whetstones) and
36% at the integer-Dhrystone-test. ubench says 20% and
2/3 of the test would be  floating-point-operations, only 1/3 of it
integer-test.

The differenz between the tests in memory is a  bit greater
ubench says my maincomputer are 186 % better than the Lap.
The-win-benchmark32-test has a difference of 247%.

The differences maybe come from the os .

Bye

Klaus Groening
Duesseldorf, Germany