GFFX Vs 9700 in Doom3 alpha

[merlocka]

Digit-life's GeforceFX review includes comparisonbetween GF4, GFFX, and Radeon 9700 in the leaked Doom3 Alpha.

The data confuses me because I recall Carmack's commenting that nv30 was much slower than R300 when they were both using the ARB paths. This data certainly shows otherwise, although I'd take it with a grain of salt since it's obviously the leaked Alpha release.

 

[Deeko]

We'll see

 

[CurtCold]

Alpha, so it's nothing compared to what we are going to see as an end product. The game will run fine on either card I'm sure.

 

[bunnyfubbles]

Something seems a little backwards there with the 9700 Pro seemingly doing better than the FX at the lower resolutions than at the high. Seems like the Pro beats the FX in almost every 1024x768 situation without FSAA or AF, of which results seemed just the oposite with other games in reviews like the one here at Anandtech. Kind of bugs me there, can't really bring myself to trust those "results" because of that...

 

[Don Vito Corleone]

I think it is highly speculative to assume the leaked alpha is representative of how the final game will run - I tend to doubt it is.

 

[kylebisme]

na its more than that, it is just plain silly. most of that article is prety ludicrus though, comparing 8xaf to 16xaf and then different fsaa styles at the same sample rate just a small note on quality differences, at least there are pics i supose.

 

[MarkFahey]

Alright, who is this guy "Alpha" and what is leaking from him???

Fx or not.. You will still be better off holding on to yer 9700 Pro or those considering buying a new video card should grab the Radeon.

 

[majewski9]

OK so it wins the alpha test in DOOM 3 but the 9700 pro was winning a considerable ammount of those benches. We shouldnt even be looking at alpha benches anyways.

 

[ScrewFace]

The only problem is that the GeForce FX only has 4 pixel pipelines with 2 texture units each for a pixel fillrate of 2 Gigapixels and a texel fillrate of 4 Gigatexels.