Why are people insisting that Futuremark doesn't know their own software?

[Keysplayr]

So whats up guys?

 

[xSauronx]

link? is this from some tech site or here in the forums?

 

[Keysplayr]

Its here in the forums. A lot of folks are using Driverheavens recent article which basically says that Futuremark approved nvidia's 6800 drivers, but insists that futuremark is wrong. And I am just very curious why (I believe I know why but still) anyone would think nvidia is cheating this go round.
I spoke of the 5950/9800XT/6800 screenshots compared to one another. You can easily see the optimizations for the NV3x card, but the 9800XT and the 6800 are so close now that its hard to tell which one is which if the screenshots arent labeled. AA is totally improved which is evident from the screenshots. AF is right on par with ATI. So IQ is no longer an issue with post nv3x cards. NV had an issue with FarCry but it was not due to optimizations but a driver bug. (after all it is a beta driver).
So I want to know why. And I am not looking for some stupid FB reasoning. I want real dialogue.

 

[Blastman]

So what?s changed about NV???? In the last couple of months NV has hacked their 8AF almost down to nothing on a lot of the benches with the FX line of cards. I?d say a lot of the benchmarks using 4AA/8AF with the FX line vs R300s are about a farce now.

THG

It is nothing short of frightening, how far NVIDIA has reduced the filtering quality of the FX 5950 U over the past few months.

If ? the R420 turns out to be faster than the NV40 (and assuming NV already knows it) why wouldn?t they continue this behavior? Also, with the limited resources of Futuremark I doubt they have enough time to spend checking for all of NV hacks. Not saying NV is cheating in 3Dmark2003 yet, because I don?t think things have been proven conclusively -- but to say NV no longer has a motive to cheat because they have the NV40 simply doesn?t wash with me. NV had a better line of cards with the 4200 - 4600 series versus the 8500?s and they were cheating back then to press their advantage even farther.

EDIT.. added THG quoate.

 

[VIAN]

I believe futuremark is an idiot because they tote that there "software" shows accurate in-game performance, and the truth is far from that. Accurate in-game performance can be found in 2001SE.

I think NV really put all of their cards on the table this year. Extremely high transistor count, lower yields, massive features, high clockspeeds. And if proven, cheating. This is laying your neck in a guillotine and hoping. Which is why I would laugh if they were still slower, but I guess I could understand their frustration.

 

[Schadenfroh]

because fanboys must justify the ATI gods

 

[CaiNaM]

it's a worldwide consipiracy keys.... futuremark is obviously cheating as well. :roll:

 

[SneakyStuff]

What goes on here doesn't recleft the whole world VIAN, just because you think something doesn't mean that the next person does. REAL gamers are more interested in how the 6800 series will perform in their games, which is something concrete, and if nvidia cheats there, it's fine by me :roll:

 

[ronnn]

Ha Ha, different folks are now supporting futuremark. Driverheaven wrote an article showing some IQ issues with the 6800ultra - something Tom and others mentioned as well. If some people jump to the conclusion that Nvidia is cheating (rather than crappy drivers), well so what. :shocked: