Back to the OP's issue--have you tried dialing LLC back a notch or two? I suspect that the "extreme" setting may be causing excessive voltage excursions that might be behind the double-boot issue. I run "High" on my ASUS P8Z68 Gen 3 and have never seen the issue up to and including the current operating freq of 4.8GHz/1.39v (on a 2600K).

Back to the OP's issue

Indeed! Apologies for the thread hijack...

*shrug*If you like!

Back to the OP's issue

Apologies for the thread hijack...

This was what I ment.

Back to the OP's issue--have you tried dialing LLC back a notch or two? I suspect that the "extreme" setting may be causing excessive voltage excursions that might be behind the double-boot issue. I run "High" on my ASUS P8Z68 Gen 3 and have never seen the issue up to and including the current operating freq of 4.8GHz/1.39v (on a 2600K).

Well I've just past a 12 hour Prime95 test 1.39V for 4.7 GHz. This as low as I can get the voltage. At 1.38 I get BSOD after 8 hours. This 12 hour test was also with LLC on Ultra High. Can High cause any instability problems? What does LLC actually do?

Arjen, quite an expected voltage for 4.7, nothing unusual.There is a big debate what LLC should be used. I just say go with the one you like - higher the LLC, higher the boost when in load, basically. You gotta see what works best for you.I ran many settings for quite a while and couldn't see any real instability problems between all.

Well I've just past a 12 hour Prime95 test 1.39V for 4.7 GHz. This as low as I can get the voltage. At 1.38 I get BSOD after 8 hours. This 12 hour test was also with LLC on Ultra High. Can High cause any instability problems? What does LLC actually do?

LLC is Line Load Calibration--it is a setting on ASUS motherboards that drives the power control circuitry to add extra voltage to compensate for the drop in CPU core voltage that occurs when the CPU is loaded up (AKA voltage droop). The theory is that when starting with already high CPU core voltage, the rapid application of the LLC correction as the load is applied or removed can drive a momentary CPU voltage spike that's enough to cause problems. Quite a few people have reported issues when overclocking SB processors where the CPU will run fine for hours, but will occasionally crash when the load changes dramatically, i.e. starting or stopping a run of P95/OCCT/IBT etc. This seems consistent with the theory that voltage overshoots caused by LLC corrections during the transitions may be an issue. I think the highest/most aggressive settings of LLC were designed for guys running exotic cooling solutions to get benchies to run for their overclocking trials, not for long-term sustained overclocks.1.39v is perfectly reasonable for 4.7 GHz.

I've tried putting on High, today I've done 3 boots already, two of them were double boots. When it rebooted after the double boot, it locked during Windows startup, Srdan had the same he said.Also tried removing the power cord, didn't make a difference.EDIT: read on the net that disabling Internal PLL Overvoltage might fix the problem. I've done that now. Just wondering if it will do anything to stability? I read that it has to be enabled for extreme overclocks, just wondering if 4.7 GHz is considered an "extreme" overclock? I also tried updating the BIOS to the newest version released in Febuary 2012, version 3202, this did not fix the problem. I don't understand why the manufacturers don't fix it in a BIOS update? Perhaps it's unfixable, and it's just a bug of the P67/Z68 chipsets? (As Srdan said).