Originally Posted by Arm3nian
You said you ran 1.55 for 6months lol.
And its not my opinion its in intels datasheet for the cpu on their site. Remember the little chart? You might be running 1.55 but I doubt youre folding 24/7, I wouldnt do anything over 1.52 but play low end games and browse the internet. There is a reason intel would put 1.52v on the graph, manufacturers dont put a limit unless there is one.
That chart has nothing to do with Vcore that is VCC range.
Originally Posted by Idontcare;34296202
<- has BS in Materials Science Engineering and PhD in Chemical Physics, directly worked on process node development spanning 0.5um to 32nm at Texas Instruments
register range =/= Vcc range
Please stop equating the two. The max value for a register has nothing to do with the max safe core voltage
And? What does any of that have to do with the fact it is not Intel's official spec as falsely claimed?
VID range simply identifies what voltages can be set/measured. It has no bearing on what is safe.
For the non-technical or uninitiated, the VID is represented by 8 bits, which means a total range of 256 values (convenient, as this is also exactly 2 hexadecimal digits for a shorter representation). It goes up to 1.52 simply because of the range these bits can express and the size of the vcc steps per byte value (one of the 256 possible combinations).
The small size is necessary for better resolution (accurate representation of an analog phenomenon using digital means). Using steps of. 005, had Intel skimped on the VID transistors and only allotted 7 bits to the task, they will only be able to set/measure up to 0.88v (.005 x 126 + 0.25; all low bits are used for zero voltage, 000001 is the starting "lowest" vcore measured, hence we subtract 2 from 2^7 or 128, which is how we got 126). So they used one more bit to be able to set/measure more values, seeing as to how it would be useful to measure stuff like 1.0v and 1.2v. Of course, the addition of that one bit doubles the possible combinations, from 128 to 256, so you end up with values that are retarded but measurable simply by nature of having bits enough to represent them.
That is why the VID range goes up to 1.52v, which does not seem at all like a round figure. It's simply the value when all available bits get used up.
"Vcc_max" in the table also does not mean what some think it means. It does not mean "this is the maximum vcore ivy can take." It means "this is the max vcore that these set of bits will represent" - remember, we are trying to digitally represent something analog, so the best you can do is set bounday, unless you plan on implementing infinite resolution using an infinite number of bits (I am exaggerating a bit, but it is very impractical to say the least).
I said i ran at 95c for 6months not 1.55v ?Edited by Hokies83 - 12/10/12 at 1:46am