While it's nice that many owners of 4-series and 6-series cards say they have no problems, it would be more helpful to read posts by owners of 5-series cards that have had problems and found solutions.
I can say that I get screen corruption (pink lines) on my desktop almost every time I upgrade to a new AMD driver release. This is usually after a cold boot and zero game-playing or video-playing. The pink lines go away after another restart, but to me, this is a sure sign of a driver problem since it's actually brought on by installing the drivers. In fact, I've never had that issue during gaming. Since November, however, I have had frequent driver crashes during game-playing and video-playing, during which the screen goes blank for about 20 seconds and a message indicates that the ATI driver has crashed and rebooted.
I've owned the card for almost a year but never had that problem until November. I guess it could be a failing card, but it's more likely a driver-related issue, as I don't have any graphical corruption in games, ever.
Glad I am not the only one, sounds like the exact issues I am having.
I know others have raised the question of if it is drivers or hardware, I am going to stick with drivers being the issue. there are many reasons but mostly is experience. Working with electronics at work, of the past 100 problems we have had, 99 have been a software issue, 1 was due to hardware (bad chip). looking over several hundred parts, that ratio only gets worse for software. One of the biggest reasons is the push for faster releases and lack of good regression testing. Too many times software engineers write off single failures as a fluke, what I have found through significant testing is that this is a sign of a larger problem, and while is sucks to try and diagnose we need to understand why if failed (or at least give due diligence to trying many things) before saying is was just a fluke. The work load is high (I understand why software engineers usually ignore it) but when reliability is key (and it should be for almost everything) the time must be spent to understand and diagnose.
From the hardware side, the two biggest causes of failure are temperature and power. Every 10C rise in temp the life of a device is cut in half. not only that, but the timing changes the series resistance goes up, which causes higher power dissipation, which causes higher temp (viscous cycle). This puts stress on all of the components and can lead to many problems (higher BER being the number one that occurs). Power is equally important, Vth is constantly getting smaller meaning even minor variations in voltage can cause problems. The other one that really sticks out is current transients due to load changes.
I did upgrade Power Supplies to ensure power was not an issue, it helped with some issues, but the video ones still remain. I also turned the fan on 100% (seeing it operate at 65C was very disturbing) and the card now holds 35C, a safe temp.
I don't want to trivialize the work to create good drivers, but when I get a 100+MB driver download, either someone is not taking care to write efficient code, or someone better make damn sure it is tested thoroughly. the number of people with issues tells me that both options are not being met.