[jwhat]

If you have an FTP I can send the game files to you, as I already have them in a .tar format.

What we really need is a central place where all of our files can be safely hosted! It would certainly make things easier for everyone!!!!

Hi Irinikus,

I have not found what the exact cause of problem is, but quake3 is now running ok.

Thank you for providing fix.

Here are some summary results (same graphics setting as your test).

Fuel with 600 MHz / 3 GB RAM / V12 @ 1280x1024

Quake 3 - 1.35 (nekoware current) : 1260 frames 53.9 seconds 23.4 fps 19.0/42.7/82.0/8.4 ms
Quake 3 - 1.36 (nekoware beta): 1260 frames 53.9 seconds 23.4 fps 20.0/42.8/83.0/8.4 ms

Octane2 with 2 x 600 MHz / 8 GB RAM / V12 with DCD @ 1280x1024

Quake 3 - 1.35 (nekoware current): 1260 frames 70.1 seconds 18.0 fps 30.0/55.6/112.0/11/8 ms

So for Quake3 the Nekoware Current & Beta version both worked and delivered same FPS.
And Fuel with single 600 MHz performed better then Octane2 with 2 x 600 MHz (both with V12).
I am wondering if the DCD factors into this as it impacts the amount of texture memory available.
NOTE: To run the Quake 3 - 1.36 (Nekoware Beta) version I had to copy the /usr/nekoware/bin/quake3 executable into the /usr/nekoware/games/quake3 directory as per Nick's guidance, as this does not have the "/usr/nekoware/games/quake3" search path compiled into binary.

I also run Quake 2 test using the the version down loaded from irix Network archive.
Using same configuration as per Ian's benchmarks @ SGI DEPOT

setenv DECOUPLE_SWAPBUF y
timedemo 1
map demo1.dm2


Fuel @ 1280x1024
689 frames, 14.5 seconds: 47.6 fps

Octane2 @ 1280x1024
689 frames, 14.3 seconds: 48.3 fps

So for Quake 2 the Octane2 was little bit faster than Fuel... and my 600 MHz machine posted similar result to Ian's 900 MHz Fuel.
This could due to V12 fill rate being key factor as Ian points out not CPU performance.

I also ran the Quake 3 demo on O350 but results were funny so will redo later.

Happy SGI gaming.

Cheers from Oz,


John.