Originally posted by: Zebo
Originally posted by: AutomaticErik
XT beats a 6800Ultra? Somebody's dreaming.
Also need to fix this:
6800U > x800xtpe > x800xt > 6800GT > x800pro
The only people dreaming are nvidia fanboys. 1600x1200 w aa/af
Yes very interesting. Of course, there are many websites that show completely opposite results. Let's take this one for instance @ Digit-Life.com (a credible sister site of xbitlabs.com):
Certainly Digit-Life has a similarly configured system apart from A64 3200+ vs. 3400+. But it has later versions of AtI and Nvidia drivers making comparison more relevant. So let's see :
I'll compare only 1600x1200 4AA/16AF like you did:
Call of Duty - Now Nvidia wins -
1
if you ever found a review where Nvidia 6800 cards are slower than X800 cards, you know it's wacked because Ati will lose 99% of all OpenGL benchmarks
Far Cry - generally a tie, but if you want to go by pure numbers, 6800Ultra beats x800xt @ 1600x1200 4AA/16AF -
2
UT2k4 -
x800xt PE wins
But 101.8 vs. 95.2 for 6800 Ultra would probably put X800xt very close to the Ultra.
Counter strike - based on number figures alone, Nvidia 6800Ultra wins @1600x1200 4AA/16AF in Cobble, Aztec, Dust and Italy levels tested here:
Counter-Strike Performance: Part 1 In reality they most likely provide similar gaming experience.
Now I dont want to go through every single benchmark, because obviously you can always find a review showing different things. But from what I've read, the consensus is that 6800Ultra provides similar performance to X800XT. Even in the benchmarks you posted, when X800XT won, it either won by 2-5 frames which is nothing, or it won average frames, but Nvidia card had the higher minimum frame in some cases (so did X800xt really win then?). Also if one card displayed 70 frames, vs. 65 frames for the other is that really significant?
The point is I'd never consider X800xt a winner if at Halo 1600x1200 16AF it gets 52.8 frames and Nvidia 6800Ultra gets 51.0 frames. That's not a clear win -
Link.
In Perimeter, Ati dominates, but is still unable to provide any type of playable performance past 1024.
But Nvidia dominates in Doom 3 and gives you the ability to turn on HDR in Far Cry. With more doom 3 based games to follow, I wouldnt want to encounter severe performance disadvantages that ATi's lineup is facing. In contrast, for the most part, all other games are equally playable on x800xt as they are on 6800Ultra.
In my eyes, since you can get 6800GT for $380 and X800xt is around $450, that is a hefty price to ask for similar gaming, and lack of doom 3 performance (as well as future games on that engine). In the future it's highly unlikely that x800xt will deliver playable performance and 6800gt will not just like 9700, 9700pro, 9800pro, and 9800xt all suck in Doom 3 even though there are 10% improvement increases in each step from one to another. Why pay extra for a card that plays at 80 frames vs. your 70 frames, when in 2 years they'll both bottom out at 20-30 frames in newest games?
Also you are saying that X800xt is faster in 70% of the games. Even if that was true, what if the user mostly plays games like MOHAA, Serious Sam, Call of Duty, Doom 3? I think it's important that users realize that they should buy a card based on the games they play the most, not based on bragging rights. If you play Doom 3 and want Quake 4, stay clear from ATI. If you play Starcraft, there is no need for 6800ultra
However, in this case GTO just sucks. But I dont think paying anything extra over 6800GT is justified at this point. Maybe I am not an enthusiast to spend $500 on a videocard?