Drdeath Posted November 23, 2011 Posted November 23, 2011 (edited) My system is stable in Prime95 for 12+ hour runs at it's current state. I'm currently using a hyper 212+ and getting as hot as 75c on those long Prime blend runs... I plan to upgrade my cooler (not quite sure what yet, thinking the Corsair H100) before I go too much further. I previously had the multi at 46 with 1.35 vcore and I started running into trouble in Prime after a few hours (one core would fail). I have not run at 46 since I bumped my VCCIO voltage from 1.1v to 1.125v... may try that again (1.1v was recommended for 4GB of RAM in a youtube guide and I'm running 8GB of RAM). 75 is a little high for a Sandybridge chip even under load........ Keep the load temps under 65 IMO. The Hyper 212+ is not the best overclocking air cooler. If you want higher overclocks unfortunately you need to entertain better cooling. Edited November 23, 2011 by Drdeath Share this post Link to post Share on other sites More sharing options...
Brutality Posted November 23, 2011 Posted November 23, 2011 (edited) Thanks, nice to hear reasoning behind suggestions. I've just got it set to auto right now, working like a champ. I just need to find a better cooler. My Hyper 212+ makes some really annoying vibrations/noises when speeding up and running at high... going to return it and get something else. It's best to avoid having voltages set to auto, to avoid crazy overvolting when overclocking. Note the value it says to the right of it and change it to that (so it's manually at that voltage, instead of auto). Good idea to refer to the motherboard manual and keep the voltages in the safe range. On my p8p67 pro rev 3.0 it's safe to stay in the yellow colour code. When you hit the temperature wall (good idea to stay below 72.4*C), you can try to decrease secondary voltages (VCCIO/VCCSA/PLL/PCH); What I did was: Set VCCSA to 0.935v, set VCCSA to recommended voltage for your dimms (don't know for 1600MHz, like I said it was 1.12-1.14 for 2133MHZ, so I put mine to 1.1375), set PCH voltage to 1.05. Now it's time to see how low your PLL voltage can go without losing stability (BTW: Only enable PLL Overvoltage when going 4.7GHz or higher, no need lesser frequencies on SB). So drop your PLL in small increments and test for stability until you reach instability.. Then put it back to when it was stable. At this point you can also try to drop your vCore (often a low PLL will allow you to become stable at a lower vCore, thus causing lower temps - credit to Nuclear for this one) Having found your lowest stable secondary voltages you can now try to get it stable at a higher frequency with lower temps =) In case of BSODs, here is a nice list of them and their cause/resolution; 0x124 = add/remove vcore or QPI/VTT voltage (usually Vcore, once it was QPI/VTT) 0x101 = add more vcore 0x50 = RAM timings/Frequency add DDR3 voltage or add QPI/VTT 0x1E = add more vcore 0x3B = add more vcore 0xD1 = add QPI/VTT voltage 0x9C = QPI/VTT most likely, but increasing vcore has helped in some instances 0X109 = add DDR3 voltage 0x0A = add QPI/VTT voltage 0x116 = Low IOH (NB) voltage, GPU issue (most common when running multi-GPU/overclocking GPU) 0x7E = Corrupted OS file, possibly from overclocking. Run sfc /scannow and chkdsk /r Edited November 23, 2011 by Brutality Share this post Link to post Share on other sites More sharing options...
Uniwarking Posted November 28, 2011 Posted November 28, 2011 I have a corsair h100 in the mail, should be here on Wednesday. Hopefully it will cool a bit better than my 212+, then I can look at upping the oc a bit more. Thanks for all of he advice! Share this post Link to post Share on other sites More sharing options...
Drdeath Posted November 28, 2011 Posted November 28, 2011 (edited) It's best to avoid having voltages set to auto, to avoid crazy overvolting when overclocking. Note the value it says to the right of it and change it to that (so it's manually at that voltage, instead of auto). Good idea to refer to the motherboard manual and keep the voltages in the safe range. On my p8p67 pro rev 3.0 it's safe to stay in the yellow colour code. When you hit the temperature wall (good idea to stay below 72.4*C), you can try to decrease secondary voltages (VCCIO/VCCSA/PLL/PCH); What I did was: Set VCCSA to 0.935v, set VCCSA to recommended voltage for your dimms (don't know for 1600MHz, like I said it was 1.12-1.14 for 2133MHZ, so I put mine to 1.1375), set PCH voltage to 1.05. Now it's time to see how low your PLL voltage can go without losing stability (BTW: Only enable PLL Overvoltage when going 4.7GHz or higher, no need lesser frequencies on SB). So drop your PLL in small increments and test for stability until you reach instability.. Then put it back to when it was stable. At this point you can also try to drop your vCore (often a low PLL will allow you to become stable at a lower vCore, thus causing lower temps - credit to Nuclear for this one) Having found your lowest stable secondary voltages you can now try to get it stable at a higher frequency with lower temps =) In case of BSODs, here is a nice list of them and their cause/resolution; 0x124 = add/remove vcore or QPI/VTT voltage (usually Vcore, once it was QPI/VTT) 0x101 = add more vcore 0x50 = RAM timings/Frequency add DDR3 voltage or add QPI/VTT 0x1E = add more vcore 0x3B = add more vcore 0xD1 = add QPI/VTT voltage 0x9C = QPI/VTT most likely, but increasing vcore has helped in some instances 0X109 = add DDR3 voltage 0x0A = add QPI/VTT voltage 0x116 = Low IOH (NB) voltage, GPU issue (most common when running multi-GPU/overclocking GPU) 0x7E = Corrupted OS file, possibly from overclocking. Run sfc /scannow and chkdsk /r With P67 and Z68 boards this is not always true. In the past, when the board took over, it would sometimes use crazy voltages but not with this platform. If the chip overclocks too high for the cooling, the bios' usually give a manual multi adjustment to clock the CPU acordingly but the user has to play with it. If the user wants: after the board overclocks, he may fine tune it and find a slightly lower voltage/ combo but the board voltage for each overclock(multi) should be pretty darn close to what is needed. These boards are much smarter than in the past. Turbo is a great way to get somewhat close to the proper voltages with P67 and Z68 boards. Also with good cooling, voltages well over 1.4V can be safely used as long as temps stay 60ish under load...... Edited November 28, 2011 by Drdeath Share this post Link to post Share on other sites More sharing options...
Recommended Posts
Please sign in to comment
You will be able to leave a comment after signing in
Sign In Now