Originally Posted by Visceral
I don't think they are "fake" I think they are stories of normal failure, etc that are being pinned on something that probably isn't the root cause. There's just not enough data, at all, to support this. Kepler is different, but not *that* different. The laws of physics still apply to the silicone in them as much as they do with any other card, past of present. the amount of voltage just isn't that great, it would mean Nvidia was cutting it insanely close if that small amount of voltage was killing these cards.
There is conclusive evidence that extreme over voltage causes electromigration, this is a fact with all silicon. I've had more than 1 cpu die from extreme overvoltage and it wasn't a heat related issue. However, as I mentioned earlier most of us are upgrading within a year or two anyway so its not a big deal - as long as you don't go stupid high on voltage you should be okay. As I mentioned earlier i'd keep 24/7 voltage at 1.25V or lower, 1.3V-1.4V 24/7 is just begging for problems. I'd only use super high voltage for benchmark runs, its insane to stay that high (1.35V+) 24/7.
So, lets just drop it shall we? This is all pointless to discuss. Electromigration should be common sense, but this is OCN. We're all upgrading sooner than typical users anyway, so who cares. I can't see something like 1.25V killing a card within 2 years, but 1.35V ? Maybe, who knows. Lets keep the discussion on BIOS' going instead of this! Dead horse has been beaten!Edited by xoleras - 8/8/12 at 1:02pm