Recent tests made by online magazine Chip.de reveal that Forceware beta drivers v.61.11 that were supplied by NVIDIA to compare against Radeon X800s during this first week of reviews, carry some game-specific optimizations, the german testers noticed a sharp decrease on average FPS when they renamed the executable file of the popular first person shooter Far Cry (example: "Fartcry.exe").

Unfortunately we have no GeForce 6800 to verify this information or to show any image quality comparison (optimized vs. unoptimized) but most likely more on the subject will come up very soon. We had heard already from various reviewers that NVIDIA used a "brilinear" hybrid filtering mode that improved performance by a small margin by sacrificing some quality.

This may seem to be an endless debate, whether you will accept game-specific performance optimizations for a rather small loss in graphics quality, or should manufacturers retain "virgin" settings with the option of turning on these optimizations?
I have a very clear answer for that, now the questions would be, should NVIDIA (or any other manufacturer) be blamed for such practices?

does this happen only with geforce 6800? What about the other FX cards?

Damn Nvidia strikes again with a bad point for them. Will they ever learn?

when people read reviews, they assume that the frames-per-second a particular card gets in a certain game or test is comparable with another card, everything else is equal. that means pretty much the same IQ, etc.

oh well, there is no industry regulation on video cards that forces a company to let the user know the details about a certain video card.

But there's a whole community behind it and that's something we should be proud of .

Indeed...

I think there are several sites working on testing the errors in the 61.11 drivers to see what really is going on...

Will be very interesting to see if we'll have a replay of what happened last year, or if it's actually a bug in the game... Since DX9.0c isn't out yet, the NV40 isn't using ShaderModel 3.0, but using the old path. This could lead to problems for the card, and nvidia might have chosen to detect the app and correct what is going on/use shader3.0...

And it starts all over again.....
Seems like there really needs to be some point of standard so all this stuff wont happen over and over.