This is a summary of the last messages. The table is sorted by sends speed.

CPU       bcodes    sends     Details
PM 1.4    228M     6.9M     Daniel Vainsencher's IBM R40 Pentium M Debian
P4 3.0     120M     6.4M     Dean Swan's Dell Win XP Pro
G5 2.0     148M     6.1M     Ken Brown's dual G5 Mac
PM 2.0   178M      6.0M     David Shaffer's Dell Inspiron 8500 Gentoo
P4 2.4     234M     5.4M      Juan Vuletich's P4 HT desktop WinXP
G4 1.2      93M     4.1M      Dean Swan's iBook
P3 1.1     119M     3.5M      Juan Vuletich's IBM T23 notebook WinXP
P3 1.0     102M     3.0M      Jon Hylands' Dell Inspiron 8100 WinXP Home
Drn.94       94M     2.8M     Juan Vuletich's old AMD Duron 0.945GHz WinXP

Some observations:
- Daniel's Pentium M is the overall winner. The (bytecode and send) speed to
clock
ratio is much better than other Pentiums and all Macs. However, David's
Pentium M is a lot slower.
- Dean's P4 performs very bad on bytecodes/secs when compared with other
P4s. Some problem with Dell machines?
- On Macs, speed (bytecode and send) seems proportional to clock.
- On PCs, bytecode speed seems proportional to clock.
- On PCs, send speed is not proportional to clock. P4s are slower than
expected.
- Mac's bytecodes to clock ratio are worse than P3's and P4's. But their
sends to clock ratio is better.

What does all this means? I don't know.

Juan Vuletich
----- Original Message ----- 
From: "Milan Zimmermann" <milan.zimmermann at sympatico.ca>
To: <squeak-dev at lists.squeakfoundation.org>
Sent: Thursday, October 06, 2005 1:38 AM
Subject: Re: OFF Topics :: PC-laptop or Powerbook


> On October 5, 2005 20:26, Dean_Swan at Mitel.COM wrote:
>> >I'm using an IBM R40 (Pentium M 1.4Ghz) with Debian and Squeak. Squeak
>> >
>> >For what its worth,
>> >
>> >0 tinyBenchmarks ->
>> >  '232304900 bytecodes/sec; 7018937 sends/sec'
>> >   '225352112 bytecodes/sec; 7028070 sends/sec'
>> >  '230838593 bytecodes/sec; 6830576 sends/sec'
>> >
>> >(hmm, that seems almost twice as fast at bytecodes/sec as the Dell
>> >reported, interesting)
>>
>> That is interesting.  It is also interesting that the sends/sec are
>> almost
>> the same
>> as the 3.0 GHz P4.
>>
>>  Can anybody explain these differences?  Could Linux vs. XP have anything
>> to do with it?
>
> Apart from there may be some difference between the Linux vs. XP VM,  this
> is
> likely because Pentium 4 and Pentium M are comppletely different
> architectures, Pentium 4's IPC is uselessly low. In the article linked
> below,
> the benchmark shows that P4 3.2Ghz (Prescott, 3200/133) is roughly
> equivalent
> to Pentium-M 1.8 Ghz (Pentium-M 735 Dothan, 1800/100).
>
> http://www.tomshardware.com/cpu/20050525/pentium4-10.html
>
> (there are more benchmarks in the other chapters)
> Milan
>>
>>         -Dean
>
>
>
>
> -- 
> Internal Virus Database is out-of-date.
> Checked by AVG Anti-Virus.
> Version: 7.0.344 / Virus Database: 267.10.16/83 - Release Date: 8/26/2005
>
>