Anyone noticed in the THG 9800 review... GF FX PS 2.0 Scores

http://www.tomshardware.com/graphic/20030306/radeon9800pro-25.html

With the WHQL 42.72 Candidate drivers, the GeForce FX Ultra was able to score a whopping 36.2 FPS in the pixel shader 2.0 test.

Compare to the 12 FPS of the 43.00 (which is the same speed as the original reference drivers in the old GF FX previews).

Now, compare to the nVidia "optimized" 42.67 drivers pixel shader 2.0 score of 15.5.

http://www.hardocp.com/article.html?art=NDI4LDQ=

Hmm... Is nVidia pulling performance out of their hats like some extremely optimistic nVidia fans hoped... Or is it another con?
 
i realy do hope the scores are valid but i am not going to get excited until i see real whql certified drivers.
 
I won't take any of the GF-FX scores as being valid until they have DX9 Compliant WHQL drivers. Unfortunately, I have a feeling that it'll be a long time until that happens...
 
Fearful

With recent news about incoming whql detonator,
i'm a bit fearful of what it will really be like.
What if it does FP32 only on 3DMark03 and does FP16 on games?
(though i have no idea if it's possible technically)
Then nvidia claim 3DMark03 is nothing, boasting fast gaming fps.
 
Has anyone done a PS precision benchmark, yet? Not just to measure performance, but precision itself? Something like Humus' Mandelbrot set demo could be used to see how deep one can go and still see more detail. ..
 
Back
Top