There was nothing faulty, they just did it differently which required some work on the drivers before performance was at decent levels again. The only thing that makes it look bad, is the fact that NVIDIA just does it extremely well at 4X AA because they are using their ROPs for the calculations and they specialized their design for that particular workload. This can also be seen by the fact that performance for both is on par for other levels of AA.
So ATI isn't doing something wrong, NVIDIA is just doing it extremely well.
EDIT: here is a linky to some more info on this matter, just to cover my behind![]()
Bookmarks