Come on, where on earth has that been said?
The only even vaguely related item was a demo of the "Samaritan" demo on the Unreal 3 engine. It was shown running on a single Kepler card using a new AA formulation. The first time the demo was shown (around a year ago) it was using three GTX580s.
There was NO statement that the more recently shown demo was running faster, and there was certainly NO statement that the new Kepler GPU would be faster than the tri-580 setup in any context (even in this particular demo with the new AA method). You're putting words into Nvidia's mouth, and then stating that they're wrong!
I honestly don't understand why you're so determined to bring down this GPU...
I'm not determined to bring down this GPU.
How many people used their right to free speech to **** on the 7970's parade?
As for the three 580s, the title of that article was "Nvidia Kepler beats three 580s" or something like that. And it did. With a new version of the code, in specific circumstances. Where was that clearly pointed out though? it wasn't, was it? it was left for people to think about.
I have said above that AMD are just as guilty. In fact, they sent out two cards for review and it was a vapour release. Where the hell ARE the 7850 and 7870 to buy? as the days roll past nothing is happening, and they are not available to buy.
So was that done deliberately to show what they had coming, yet wasn't anywhere near being available? of course.
I'm just calling a spade a spade. I do that with everything. Want to know how great my 7970 is? Well, not only have I wanted to cut it in half and smash it with a hammer this week but I have also made my thoughts clear on how bad I think the drivers are.
So the only thing I can be accused of is being a party pooper, right? Then good. If I thought for one moment those benchmarks were fair and just then I would be getting just as excited as others seem to be. But they are just completely biased and loaded. That is just a fact I'm afraid.
I want Kepler to be good. I want Kepler to be
truly good
I don't want Kepler to be good, if viewed from a certain angle, at a certain time of day, running a certain game and on a certain set of test equipment.
Does that seem fair to you?