My burn in has been done sloppily. I pretty much did it the way Nade told me until I started getting a lot of crashes at around 2650 the first day I really did it. It was more of a thing that I was choosing to do and bend the rules a little. I was trying to get the absolute minimum stable vcore to burn it at and I began to have problems around 2650 because 2590 is where my 1.30 needs to be upped to 1.325+. So I must have crashed and rebooted a dozen times between 2650 and my target of 2750 during the day I was initially burning-in.
So I can not say my work was well done, I never kept notes and everytime my computer crashed I rebooted as quickly as possible at the next higher voltage, burn in for a few hours and try to lower the volts back to the crash level and burn more (hopfully with no burn-in interrruption). I was eventually able to reduce and seemingly eliminate those crashes once the burn-in ran for a few hours, then it would be lowered to the crash vcore. But I feel it's more of a warming-up the computer effect as I've noticed that it tends to be more stable whenever I have it burning in for a few hours.
As a result, I've have been able to reduce vcore down as much as 0.5 to 0.6 volts settings that it would normally crash if I were to just jump in and start burning at that very unstable low voltage.
However, I can tell you this, depite my sloppy record keeping that I pretty much keep in my head, I am able to run 2750 at 1.425 volts off the bat now instead of 1.45. One example of how low I can go is if I start burn in at 2750 and 1.425, I can eventually get it down to 1.393 after it's been burning uninterrupted for a few hours. Of course after it's cooled down and restarted it just wont be burn stable at 1.40. Failes almost immediately. I do believe his process has done something to enhance the chip: I am getting more stable at the same mhz than before I started. And overall, I've been able to reduce the vcore at least 0.025+ at the same settings. I've never ran my cpu at 2850 with out it having an error in a few seconds. But the last 2 days I've been able to do it up to 40 minutes at 2850 and even a few minutes at 2900. But, I can't say that I really practiced this high frequency stability very much before the burn in. I'm just too lazy to keep notes, take time and write everything. I do feel confident that this method has for sure shaved off at least 0.025 volts and increased the length of time my cpu can be stable at those volts at higher frequencies (2750 and higher). I'm currently working at 2850 and 2900 at 1.5-1.525 and 1.550-1.575 respectively. I think I'm seeing results and am trying to push my core0 more. I am ONLY burning in core0 becuase I just can't imagine what my temps would be if I burnt both cores at the same time. I'm hitting 47-52c with that one core and 2900 hit 55c today. It was too high, had to shut it down. My other core doesn't need the burn in until core0 catches up.