Ivy Bridge-E RAM Voltage (Official Spec)?

but also because of the memory controller on board all modern cpu there is a limit that is imposed to prevent damage, in Intels case I believe it was tied with QPI if I recall, so if one is jacked then other can be pushed to its burn point.

Either way there is no "official" rated voltage but Intel did point out with Sandy-E not to go over 1.5v or it could shorten lifespan of cpu, so then you have 1.35v chips(such as the Samsung 30nm crazy overclockers) or 1.5v made by many, I would wager go with something like the Gskill or sammy low voltage(1.35v) if you really want to make sure.

Oh yeh and yes ram voltage can change, how do you think we devolt or overvolt ram to try to get more speed out of it when overclocking and such? xd yes it is ram dependant BUT some chipsets need more or less voltage to run the exact same kit this is why they have a range of +/- that rating to take this into account, so 1.5v usually is ~1.45-1.55v or thereabouts for running at rated speed and timings, in Intels case this can get tricky as yes the XMP and such has preset figures but the extras they put into their modern designs can also stipulate slight variances for timings, voltage supplied to the ram as well as to the chip/chipset that can and does cause issues for some folks, anyways.
 
@dragonstongue - Intel revised the spec for SB-E As long as you keep the IMC voltage within ~.3-.4 of the RAM voltage you should be fine.

I still would probably not go over about 1.6v on the RAM though as Intel lists the IMC voltage max as 1.2v
 
1.35v max for sandy/ivy wasn't it so should be safe 1.6 at most, still I would call it 1.5v jut to make double sure as this still gives it that +/- area no>?
 
dragonstongue said:
1.35v max for sandy/ivy wasn't it so should be safe 1.6 at most, still I would call it 1.5v jut to make double sure as this still gives it that +/- area no>?
Click to expand...

Not sure. Last time I looked at the tech docs from Intel, they had removed the specific voltage and it said to keep the IMC voltage within .3 to .4 volts of the RAM voltage. It actually might have said .5.
 
In specification was 1.5 RAM 1.5V, 1.35 RAM 1.35 V. And then there were some numbers for other stuff. Like VTT should stay at 1 V.

I actually don't think these numbers are completely valid. Intel might do the same mistakes as professional users, and they can set some numbers, which they can't test well, low just to be on the safe side.
 
ummmmm, not exactly Intel has to and does test x amount of chips off of the line for warranty and safety purposes especially seeing as system integrators will be using them, dell, hp etc also needs 100% certainty they will operate within a given range for temperatures and so forth. Intel, AMD, Nvidia and the "makers" have access to and use mucho advanced tools to help them do this. Of course there is a +/- side of things as nothing is perfect however their level of understanding of their product in regards to say a chip will last x amount of time at x temperature with x voltage in a given range is as close to it as we are going to get, they have been doing this for many years compared to the vast majority of users.

and where is this spec you are quoting?
 
It would be nice if they would release these precise and accurate data into general public.

I use 4th-gen-core-i7-lga2011-datasheet-vol-1.pdf as a source of info about new Ivy-E CPUs. The important parts are tables 7-10 and 7-13?.

Yes Intel wrote a load of text about thermal limits, which is what every overclocker knows from theirs experience and doesn't need to look at some paper. In fact it looks like it was writen for people who are just reading a paper and then doing it blindly and quoting numbers from paper instead of making experiment and finding facts themselves. There is very little info about substrate limits and effects of high voltage on degradation.
 
not worth it, the way you are stating what you are saying to me is coming off as the "average" overclocker knows as much if not more on these things then Intel who for the most part built nearly every chip currently in use and generally speaking also made the vast majority of rules, templates, specs, testing etc that nearly all overclockers use, yes it would be nice to see some more info however most people don't need to know all the guts of everything a simple do not pass x voltage at x temp suffices, xmp is or was meant to keep things simplified, programmed by intel to ensure that certain thresholds are not surpassed, hell you might as well blame asus and such just as much cause they DO have access to this raw data which sometimes is locked away to the public eye for various reasons.

Anyways not worth figuring out, if Intel did a lot of their own testing and released even a bit of info, guess where the overclockers probably got the info, not from their own testing but rather the info that intel gave to review sites and the average person goes to see at least some of the info so they can learn and go from there, Intel knows this and is not going to waste time or energy trying EVERYTHING out so you do not have to. They quote the numbers that matter, not every and all possibilities.
 
You must log in or register to reply here.
Back
Top