Using Quake As A Bench Mark

[R0B]

Hay, why don't we bench mark all of the ppc's to compare their speeds. Here are some rules.
1. Have between 17-20 megs free
2. Perform a softreset before each benchmark
3. Use an autoexec.cfg file with the command to run a timedemo of demo1
4. Have no "tricks" to boost the fps. I.E. the timedemo demo1 line should be the only line in the autoexec.cfg.
5.Perform 3 tests and list each result and then the average.
6. Use the 240 x 160 res (I think that is right)
7. Keep the device pluged into external power.
8. Do not have the device conected to the comp (i.e activesync).

I think that is it. Please list results/device/processor speed bellow.

[johannpublic]

Has anyone done this yet?

I'll try it out. Not like anyone wants to know the results for an old Cassiopeia...

[Robotbeat]

[Guest]

[Dan East]

In order to play Portrait PQ supports 240x160 (PQ2 uses 240x180). It will also play in landscape at 320x240, which is a true PQ and PQ2 resolution that can be compared to desktop machines.

Dan East

[Robotbeat]

Is there a way to set the default resolution in PQ to 320x240? If so, that would probably be the best for benchmarking. That way, as Dan said, it could be directly compared to Quake performance on the desktop, as long as the same default autoconfig (except with minor, non-performance affecting entries) is used for both. That would be interesting to compare PQ to Quake on 486s (is it possible?), Pentium 75s (is there a Pentium 50?), and higher cpus, besides comparing PQ performance on various Pocket PCs.

[Dan East]

You can edit autoexec.cfg to set the appopriate vid_mode at startup:

vid_mode 2

Dan East

[Robotbeat]

Cool. I didn't know that. Thanks!