nVidia promises to blow up the bank with 270's

I assume the 560Ti 3.6x faster in Dragon Age II bar being longer then the 580's 4.5x faster bar. Either saying that 4.5 < 3.6 or 580 gets less fps even with the boost then the 560Ti

 

3.6 > 4.5 ???

Probably supposed to be 4.6, judging from the image.

 

Yeah, the graphs actually start at zero. Must be fake then

 

I also see that new drivers carrying the promise of performance increases are still struggling to beat single-card performance with SLI. Wut?

 

BlackOps SLI gets slower?

 

Personally I find it strange that 560 Ti SLI beats 580 SLI in HAWX2. Also Just Cause 2 seems to be slowed down by SLi for both cards so it's probably real (unlike the L4D2 bench for 580 where the results are most probably swapped).

 

Seems a bit like when AMD has compared Mjöllnir (aaahhhhhhhhhhhhhh the UMLAUT!) to half-a-year-old drivers at first. I mean, really, compared to the 267.59, gains in Dragon Age are in the 20-percent range - not bat anyway, but 6.3x looks better in bar-format .

 

is nvidia boasting that its drivers were formerly so bugged, one game ran as a slide show?

 

notice the examples theyve actually shown framerates for only show small gains <10%
where are the framerates to prove da2 gets between 360% and 620% gain

 

They don't want to show that DA2 was running in single digits fps

 

I wouldn't say a 300% improvement , but never the less that is quite acceptable :

Before






After
​

Before




​

After



http://www.gamegpu.ru/rpg/rollevye/...eno-s-drayverami-versii-geforce-270-beta.html
​

 

There's 3 possible choices behind such extreme performance gains:

1 - utter and complete incompetence on behalf of driver team limited performance in previous version.

2 - artificially limiting performance, deliberately slowing down to be able to show huge gains at a later date.

3 - Cheating to fake higher performance than otherwise possible if doing everything properly.

 

Doubt that was a driver problem, difference is way too drastic for that. Game was probably doing something that was friendly to AMD but caused havoc on nVidia's stuff. I'm sure they just rewrote the shader.

@Grall, lol I like how all of your options are essentially saying nVidia is corrupt or incompetent. It can't be the software doing something funky eh?

 

In Grall's defense, nv has pulled many a shenanigan. Only a fool would label them as innocent until proven guilty. Impressive numbers. Let's see if they hold up under scrutiny.

 

You forgot one thing (from your perspective anyway):
3a - artificially hampering performance in gaming evolved titles so as to be able to claim AMD deliberately crippled their code.

And seriously, it could just have been an erratic driver behaviour. It's not that a game running below expectations never happened in the past from any IHV.

They certainly don't, compared to some newer 267 drivers. Going from there, it's only a ~20% increase.

 

There are millions of lines of driver code that work perfectly fine in hundreds of applications but anomalous performance in one app is obviously due to the incompetence of the driver team? Excellent deductive reasoning there.

And speaking of fools, large multi-national companies don't play idiotic games like intentionally limiting performance of their product. There is zero benefit to doing so. Check out the DA performance review at [H] - great result for nV there aye?

 

Yeah, I admit I may have been a tad overly cynical there, but I felt suspicious of claimed gains on the order of several hundred percent. That should not be possible under normal circumstances.

NV has pulled some amazing stunts in the past (like silently replacing shaders a couple years ago), and wasn't there some controversy over NV doing a funky li'l dance behind the scenes with how the driver handled 3dmark, in order to buff the scores?

While Carsten's suggestion is indeed also a possibility, is there any actual evidence of AMD doing something like that of late? Last time AMD got caught with their hand in the cookie jar (that I, perhaps selectively, can recall), it was regarding the quack controversy which was QUITE a number of years ago now.

 

Heehee now imagine if Dragon Age 2 was a TWIMTBP title, and the performance on AMD cards vs their Nvidia counterparts would be much much worse than in other titles.

Where's the outcry about Gaming Evolved?

 

hmm , I would say that "History tends to repeat itself" :


Shift 2 behaves in the same way Shift 1 did , Bad AMD performance .