I'm not sure if this will interest anyone, but I was curious to see how my voltages compared. Here are tables for different OCs (based on GHz to nearest 100 MHz) for average voltages and also their percentile ranks. The value for n represents the number of data points (overclocks) for that particular frequency. Stdev is the standard deviation. Percentiles show the voltage for which X% of the observations occur above. So if your vcore for 4.7 GHz is 1.368, your vcore is lower than about 75% of other vcores at 4.8 GHz. The larger the sample size; the more reliable. There wasn't much data for some of them, so those aren't as reliable. Maybe this will help someone know where their chip stands (it makes my chip look fairly bad
). Just realized I didn't account for 2500k/2600k/2700k, crap. Some of the data sets may be too small to try to include that though. Any feedback is appreciated.
Originally Posted by dVeLoPe
1.40v and under GOLDEN PRO RAPEFACE Z)GMOGOMG
1.45v and under GOOD/DECENT BUT NOTHING SPECIAL
1.50v and over .. TERRIBLE SELL ASAP FAIL CHIP LOL
That's really close to the statistics
As for my OC, I tried a 12 hour custom prime95 run last night, it BSOD'd (error 124) around 3-4 hours in. I tried 4.1 GHz @ 1.235 vcore (in bios), cpu-z reads 1.224 v. I was fine in custom 1344 and 1792 for 20 mins, as well as 20 pass max IBT, Should I try a little more vcore? My mobo (msi z68a-g45) doesn't have any of the other things to fix for error 124 (VTT/VCCIO or PLL, only Internal PLL overvoltage, which is currently set to auto (could enable or disable as other options)). Also, even with spread spectrum disabled, I still sit at 99.8 MHz bus speed. Is this ok, aside from giving me uglier speeds?
edit: When I went into my bios, some how the c-state options were turned off and one of my power limits was 200 instead of 250. I fixed those (enabled c-state, no limit)
ThanksEdited by Molybdenum - 1/17/12 at 10:34pm