Read the hexus reivew of GTX 590 and I enjoyed it very much thank you. Just went on to bit-tech (don't ask why), saw their review for the GTX 590 and thought I would have a peek. They have GTX 590 consistently faster than the HD 6990 OCed in every test (interestingly they have no XF or SLi setups because presumably it is impossible to skew the numbers for those in favour of the GTX 590).
But the two things that I find really strange are as follows:
Hexus has the 6990 faster than the 590 in COD:BO at 2560x1600 but bit-tech has the GTX 590 faster ... I worked out that it is because they only have AA dialled up to 4x ... why the would you not max everything out on a half a grand dual gpu GFX card!?
Second thing that is really weird is they have idle power draw for the stock HD 6990 lower than the GTX 590 but the OCed Hd 6990 has much higher IDLE power draw than the GTX 590 ... how the is that possible!? Here they simply must be making up the numbers ... if you look at the hexus review the HD 6990 stock and OCed draw an identical number of watts as you would expect at idle. I really think bit-tech must just make it up.
In summary you can get away with using specific levels of AA and AF or whatever to push a product but you simply cannot just make numbers up which they must of done for some of the results in that review.
Opinions?