Re: BenchMarrk GEforce3 under Linux

New Message Reply Date view Thread view Subject view Author view

From: Angus Dorbie (dorbie++at++sgi.com)
Date: 07/03/2001 14:45:31


It is widely accepted that unless you are interested in the GeForce3
feature enhancements there is no significant improvement in getting one
over the GeForce2 Ultra. Carmack wrote an interesting .plan on this
topic.

These features include high res AA, 4 simultaneous multi-textures, 7
general register combiners + one final, programmable vertex & pixel
shaders.

If you are buying a GeForce3 simply for a speed hike with an app that
runs well on a GeForce2 then you are buying it for the wrong reason.
It's difficult to disagree with Carmack, who concluded that any graphics
programmer worth his salt should get a GeForce3.

Use those 4 textures, Performer supports them, or play with the register
combiners, high res aa, & other goodies.

Cheers,ANgus.

> VSM wrote:
>
> I make a little comparaison between the new GEForce 3 and the
> GEforce2 ultra, the result is not amazing.
> on Red Hat 7.0 - linux kernel 2.2.18 - XFree86 4.01a and Performer
> 2.3.1 - 10 - NVIDIA driver 1.0-1251
>
> with perfly, and only one light source on a big DB:
>
> With a GL-FSAA-MODE 1 on 1024x768x16 : We win less than 20% of Drawing
> , With a GL-FSAA-MODE 4 or 5 we lost performances.
>
> On 1280x1024x16 , I have not a GeForce2 benchmark but I think we win
> about 30%.
>
> About the image Quality, I don't find a big evolution about
> antialiasing ( in all mode), I'm very disapointed by the results.
>
> Maybe I've a driver problem, or a bad optimization.
>
> If someone has another benchs , or more informations about
> optimisations.....
>
> Regards.
>
> Yoël
>

-- 
For Performer+OpenGL tutorials http://www.dorbie.com/

"Whenever there is a conflict between human rights and property rights, human rights must prevail." --Abraham Lincoln


New Message Reply Date view Thread view Subject view Author view

This archive was generated by hypermail 2b29 : Tue Jul 03 2001 - 14:43:12 PDT

This message has been cleansed for anti-spam protection. Replace '++at++' in any mail addresses with the '@' symbol.