I've been running a stable OC of 175 base clock, turbo on, 1400mhz CL7 RAM 1.65V for sometime now. Not too long ago I decided to give it a go with a slightly higher OC. I managed to get it nice and stable at 178 base clock, turbo on, 1426mhz CL7 RAM but it needed 1.68V to be stable.
Suddenly the other day, STALKER took forever to load and had a stuttering issue that was occuring when I had 4gb RAM... I alt tabbed and opened performance monitor window, and it showed only 2.8gb RAM installed! It was working fine for a couple weeks and I didn't change anything, so what happened? I tried taking out 2 RAM sticks but it still showed 2.8gb. I then changed my OC back to the 175 base clock 1400mhz at 1.65V and it showed up 4gb again, plugged in the other 2 sticks and voila, back to 8gb no worries.
So, was the RAM simply not able to handle 1.68V? Was it maybe that I needed some more voltage elsewhere (IMC/VTT?) And why was it only showing 2.8gb? I mean that's 1.5 sticks of RAM... wierd amount. Could it have been that the PSU couldn't supply enough voltage (although it's quality and has worked great so far - also, it wasn't really under load)
Also the RAM has never felt hot and the case is very cool inside...
ASUS P7P55D PRO
Zalman CNPS10X Extreme
4x2gb Patriot Viper II Sector 5 RAM (1600mhz CL8 1.65V spec)
two Sapphire 5850s at 875/1205 in CF
two Seagate 7200.12 500GB HDDs in RAID 0
EVGA GT 240 for PhysX
AuzenTech Bravura 7.1 sound card
Antec 902 case (6 case fans)
Corsair 750TX PSU
Well, they do have an "extreme" XMP setting for 1.7V although it takes even more than that in reality (I didn't try over 1.7 at the time) which would put them at 1600mhz CL7. I read someone got it working at 1.75V... but anyway, thanks I guess that answers it. Any idea why it would have worked fine for a while and then suddenly stop?
Intel's 'recommended' threshold is 1.65v. This doesn't mean that you can't get your computer to boot when you volt higher than the limit, but if any damage occurs, you assume the liability and void your warranty with the product.
I'm not surprised that somebody figured out a way to raise the voltage of the RAM above the 1.65v limit, but you must also consider what that particular build was made with.
As to why it would work for some time, then suddenly stop, I have no idea.