I'm no expert, but with mtex dose more TMUs help?
If so you should try with the patch, I have had 6 enabled with no problems or use.
Also how can I help get this patch commited? I can run test programs.
The patch should be in list arcives dated...
Sat, 3 Jan 2004 00:30:35 +0100
Sun, 4 Jan 2004 14:50:34 +0100
--- Chris Ison <wildcode@...> wrote:
> I have found something disterbing about DRI and was hoping you would be able to give a solution.
> It turns out, through recent QuakeForge changes, the DRI is faster in multipass rendering than
> with multitexture (mtex) rendering. The same method, and with no alteration to the code, with
> nVidia drivers in linux shows mtex is faster.
> It was also shown that mtex rendering was faster in windows with the same code than multipass
> using ATI catalyst drivers, and for nVidia, with the difference being from 10 to 30% performance
> improvement overall with multitexture compaired with multipass rendering. This test was done
> showing consistant results with several machines, multiple runs, and same process runs
> (switching mtex of and on in the same process). This was also seen in linux for nVidia.
> Can you please suggest anything that could improve performance in DRI so that multitexture is
> faster than multipass as seen with nVidia in linux, and nVidia and ATI in windows.
> Thanks in advance
> Chris Ison
Do you Yahoo!?
Yahoo! Finance: Get your refund fast by filing online.