GF FX/R9700 and Doom III

[HendrixFan]

Ive seen posts here and on other message boards stating that the GF FX will run Doom III faster than the R9700. They say the fillrate will allow it to push the very high polygon count that Doom III has.

However, from what Ive seen of Doom III, the polygon count is actually very low, especially when compared to the new Unreal engine. Its my understanding that Doom III doesnt use a high polygon count, but instead uses bump mapping, which seems to make sense based on its gameplay. The models look to have very low polygons to me.

Does anyone know whether or not Doom III uses a high polygon count? Are there any tests that just utitlize bump mapping, so we can see a comparison of the two cards?

 

[VBboy]

http://www.gamershell.com/game/FPS_Doom_3.shtml

It seems to have a very high polygonal count on the monsters, and lower count in the static game objects. Hey, you think those rounded surfaces come for free?

That's all I can say. If you're thinking about buying a videocard right now, don't. Wait for official benchmarks.

G'luck.

 

[Whitedog]

I remember them saying when they originally created the monsters, they were around 800,000 poly's, but them they scaled them way down once they were done. I never heard and comments on the poly count of the scaled down models.

As long as I can get my hands on a shotgun to shove down their throat, I don't care.

 

[TheWart]

I too was under the impression that Doom III was going the way of lower polies and better textures and Unreal Warfare engine was pretty much the opposite.

Has anyone seen any benchmarks comparing the GFX and 9700 on Doom III yet?

 

[merlocka]

Originally posted by: TheWart
I too was under the impression that Doom III was going the way of lower polies and better textures and Unreal Warfare engine was pretty much the opposite.

Has anyone seen any benchmarks comparing the GFX and 9700 on Doom III yet?

Gosh, considering the reaction to the GFFX on beta drivers with a cooling system which hasn't been finalized I would LOVE to hear the general reaction to benchmarks of the GFFX on an ill-released leaked alpha of a game which won't be released for 6 months.

 

[magomago]

Originally posted by: merlocka

Originally posted by: TheWart
I too was under the impression that Doom III was going the way of lower polies and better textures and Unreal Warfare engine was pretty much the opposite.

Has anyone seen any benchmarks comparing the GFX and 9700 on Doom III yet?

Gosh, considering the reaction to the GFFX on beta drivers with a cooling system which hasn't been finalized I would LOVE to hear the general reaction to benchmarks of the GFFX on an ill-released leaked alpha of a game which won't be released for 6 months.

Its Nvidias fault for releasing "beta" drivers. They have had WAY to much time to work on them and could ahve done so while they waited for the final yields to come out better.

It is like saying "It isn't ATis fault that when the radeon was first released it had horrible drivers!"

 

[CurtCold]

As long as I can get my hands on a shotgun to shove down their throat, I don't care.

LoL That would be good in a sig!

I can't wait for DOOM3, that's going to be a good excuse for me to get a 9500 or 9700 depending on the price at the time!

 

[EdipisReks]

from the screen shots i've seen, the Unreal II engine has a higher polygon count and high quality textures.

 

[TheWart]

Originally posted by: EdipisReks
from the screen shots i've seen, the Unreal II engine has a higher polygon count and high quality textures.

Oh, I did not mean that UT2k3 does not have high-quality textures, I just meant that IIRC correctly, Doom was using more bumpmapped and other super-high quality texture features.

 

[codehack2]

Originally posted by: HendrixFan
Ive seen posts here and on other message boards stating that the GF FX will run Doom III faster than the R9700. They say the fillrate will allow it to push the very high polygon count that Doom III has.

Nit picking here, but fillrate is really independant of how many polys a card will push. Poly throughput is dependant on the vertex engine/shaders and triangle setup engine for non-TL enabled games. Fillrate relates to painting those polygons, and the numerous pixel effects (bumpmapping, fog, specular lighting, etc).

CH2

 

[merlocka]

Originally posted by: magomago

Originally posted by: merlocka

Originally posted by: TheWart
I too was under the impression that Doom III was going the way of lower polies and better textures and Unreal Warfare engine was pretty much the opposite.

Has anyone seen any benchmarks comparing the GFX and 9700 on Doom III yet?

Gosh, considering the reaction to the GFFX on beta drivers with a cooling system which hasn't been finalized I would LOVE to hear the general reaction to benchmarks of the GFFX on an ill-released leaked alpha of a game which won't be released for 6 months.

Its Nvidias fault for releasing "beta" drivers. They have had WAY to much time to work on them and could ahve done so while they waited for the final yields to come out better.

It is like saying "It isn't ATis fault that when the radeon was first released it had horrible drivers!"

err, umm, no.

First of all, nVidia hasn't "released" any drivers. They haven't "released" the card yet... mmkay?

The performance of the GFFX hardware and software is what it is. The general population seems completely under whelmed by the performance, which isn't surprising considering the hype surrounding the card and the performance bar raised by ATI. Based on the paper specifications, I expected much higher performance in non-bandwidth limited benchmarks.

What I do find funny is the concern over the cooling of the card, despite the fact that nVidia has already stated that it's completely up to the AIB mfg to specify the cooling, and that they were aware of the noise issue of the card and that it will be reduced considerably in the final reference design.

Based on the fact that the general population can't even figure that out, do you think that benchmarks of an alpha Doom3 (which, at that point in the development, more time was probably spent making sure it didn't crash during the demo than making sure it would run on anything other than R300) on nv30 vs R300 would proffer any valuable data? And how misconstrued would that data be?