Performance quirk on old Thinkpad v.s. my AMD64 PC

[Guest Kal]

The two computers involved;

IBM Thinkpad R40 (Type 2681) (Roughly 2 years old iirc)

Pentium 4 M 1.8Ghz

256MB

ATI Mobility w/16MB

(OS X compatible network/audio work fine)

 

AMD64 3200+ (Winchester core)

1GB

Geforce FX5900XT 128MB

(Again, compatible network and audio - though audio is pitched up)

 

The story so far;

Okay, both PC and laptop are running IDENTICAL installs of OS X. Same exact images used, same exact commands used, same 0.5 patches, same resolution, colour depth etc. As near as damned identical as I can make them. Both are running SSE2, the laptop uses VESA 2.0 and the PC uses VESA 3.0. In every possible way the AMD machine should kick the laptop to pieces in terms of performance, right? Much faster processor, stacks more memory and a far superior video card etc.

 

A simple test shows me otherwise. Namely, the standard "Flurry" screensaver. Running on the laptop, it will work full-screen at 100% speed that it should. No dropped frames etc. It is perfect.

On the AMD64 PC it is a completely different story. The same screensaver runs absurdly slow. Jerky, dropped frames, whatever you want to call it; It runs like {censored} no matter what tweaks I try.

Using XBench, the ONLY thing the laptop outperforms on is the opengl/quartz tests. Quartz etc are most definitely not enabled on either machine, and no kexts appear to be loaded for either machine either.

 

Can anyone reason why on earth this could happen?

[KENNYB]

The two computers involved;

IBM Thinkpad R40 (Type 2681) (Roughly 2 years old iirc)

Pentium 4 M 1.8Ghz

256MB

ATI Mobility w/16MB

(OS X compatible network/audio work fine)

 

AMD64 3200+ (Winchester core)

1GB

Geforce FX5900XT 128MB

(Again, compatible network and audio - though audio is pitched up)

 

The story so far;

Okay, both PC and laptop are running IDENTICAL installs of OS X. Same exact images used, same exact commands used, same 0.5 patches, same resolution, colour depth etc. As near as damned identical as I can make them. Both are running SSE2, the laptop uses VESA 2.0 and the PC uses VESA 3.0. In every possible way the AMD machine should kick the laptop to pieces in terms of performance, right? Much faster processor, stacks more memory and a far superior video card etc.

 

A simple test shows me otherwise. Namely, the standard "Flurry" screensaver. Running on the laptop, it will work full-screen at 100% speed that it should. No dropped frames etc. It is perfect.

On the AMD64 PC it is a completely different story. The same screensaver runs absurdly slow. Jerky, dropped frames, whatever you want to call it; It runs like {censored} no matter what tweaks I try.

Using XBench, the ONLY thing the laptop outperforms on is the opengl/quartz tests. Quartz etc are most definitely not enabled on either machine, and no kexts appear to be loaded for either machine either.

 

Can anyone reason why on earth this could happen?

<{POST_SNAPBACK}>

 

 

Have you read any threads? By default nforce boards do not have proper ATA kexts loaded hence the atrocious disk transfer rate. But you can change this and get respectable speed.

OTOH, intel boards do have the proper kext loaded, for the most part. OGL may work on some ATI cards, while Nvidia peeps are SOL for the forseeable future.

And of course QE is not enabled.

[Guest Kal]

I have read plenty of threads. Neither the PC or the laptop use an NForce board. The laptop is an old Intel-based Thinkpad, and the PC uses an Asus AV8 board. Still, something I can look into now...

[KENNYB]

I have read plenty of threads. Neither the PC or the laptop use an NForce board. The laptop is an old Intel-based Thinkpad, and the PC uses an Asus AV8 board. Still, something I can look into now...

<{POST_SNAPBACK}>

 

Well, yeah the thinkpad, like my T-40 has an intel chipset. Plenty of threads have discussed QE not working or non-intel ATA kexts not loading by default.