. . . then make the mainboard supply as much voltage as the CPU is calling for in an overclock situation? Or, does AUTO mean the BIOS reads the correct vCore that is specified for the particular processor, and sets it accordingly? For example, a Q9650 stock voltage is I believe supposed to be 1.25v, so if set at AUTO, 1.25v is supplied to the CPU?I ask this because I thought for sure it was the latter of these two explanations. But when I looked at CORE VOLTAGE in lavalys, it shows when I have my DRAM running at 1640Mhz vCore is at about 1.36v, suggesting the BIOS setting of AUTO delivers MORE than the specified stock voltage to the CPU. I further corroborates this by going back and manually setting vCore to 1.25v while at 1640Mhz. The machine would no longer boot. Seems a little troubling leaving it on AUTO, especially if you are not looking to see what vCore your actually supplying.And, what is the role of CPU VID? I see this is a lot lower than vCore.Noel

Hi Noel, You found the same thing that most of us have. Leaving the Vcore in Auto scares the bejeezus out of me. I was testing a first gen P67 with a 2500K, and found Vcore at 1.5v with CPUz. Scary.If I am manually clocking the CPU, RAM or Northie, I set it manually to what auto gives me, then lower it until I get a failure, then bump it back up one or two increments.The Vid is basically the starting point for Vcore when in Auto...Don

Hi Noel, You found the same thing that most of us have. Leaving the Vcore in Auto scares the bejeezus out of me. I was testing a first gen P67 with a 2500K, and found Vcore at 1.5v with CPUz. Scary.If I am manually clocking the CPU, RAM or Northie, I set it manually to what auto gives me, then lower it until I get a failure, then bump it back up one or two increments.The Vid is basically the starting point for Vcore when in Auto...Don

That's a nice synopsis thank you Don. I kinda was thinking along those same lines last evening when I noticed. I was running at 3.8Ghz on the new Q9650 and sensed vCore on AUTO was at about 1.35, so I backed down to 1.32 and totally stable, so that's good, I'll shoot for lower vCore later today. I can now attest overclocking can and does kill processors or at least I believe so, and I had quite good cooling I believe (core temps usually below 55C or so, on occasion as high as 64. When I saw that and read Intel's disclaimer on temps I decided to back off to 3.88Gz most of the time on my now dead QX9650. And I did have my vCore and other parameters set manually. Can I also assume AUTO for other parameters such as NB and DIMM voltage is the same, ie it's best to set them manually?

. . . then make the mainboard supply as much voltage as the CPU is calling for in an overclock situation? Or, does AUTO mean the BIOS reads the correct vCore that is specified for the particular processor, and sets it accordingly? For example, a Q9650 stock voltage is I believe supposed to be 1.25v, so if set at AUTO, 1.25v is supplied to the CPU?I ask this because I thought for sure it was the latter of these two explanations. But when I looked at CORE VOLTAGE in lavalys, it shows when I have my DRAM running at 1640Mhz vCore is at about 1.36v, suggesting the BIOS setting of AUTO delivers MORE than the specified stock voltage to the CPU. I further corroborates this by going back and manually setting vCore to 1.25v while at 1640Mhz. The machine would no longer boot. Seems a little troubling leaving it on AUTO, especially if you are not looking to see what vCore your actually supplying.And, what is the role of CPU VID? I see this is a lot lower than vCore.Noel

Hi, generally it is recommended to manually set all parameters when overclocking and not leave it to "auto." I say this although right now I have my Vcore set to auto but I have an eye on it.I am not a 100% certain how voltage is determined in auto but think it is a figment of the CMOS based on VID and frequency the processor is ran at and current draw.I do know that VID and Vcore are not the same and a lot of people seem to misunderstand this difference. Voltage Identification (VID) is a unique voltage assigned to the processor as part of the binning process at point of manufacture. A chip design is assigned a minimum and maximum voltage specification that it is expected to fall into. In example my chip has a maximum VID of 1.52v but could have an assigned range of .98-1.52v My chip has a VID of 1.365v (I think off the top of my head). This means my chip passed the thermal design tests at a maximum of 1.365v It does not mean I can crank Vcore up to 1.52v as a matter of fact anything over VID of 1.365v and I would be out of thermal design for my chip I would then need to do something to compensate. The next person’s 2600 chip may have a VID of 1.356 it is unique to each chip but all 2600k chips must be in the range of the lowest VID spec (I do not know what that is) but the maximum is published and is 1.52vA lot of folks are taking that published VID of 1.52v and running around and telling everybody it is ok to crank Vcore to 1.52v ooh very bad.Another common misconception is that you can crank all the volts to the chip that you want as long as you keep the temps down, ooh very bad. Transistors are sensitive to voltage; while keeping the heat in check may decrease certain parts of the processors degradation it won’t help the transistors when you overvolt them. The internet can be a very misleading place, unless you personally know the author, people need to be very careful with what they read (present company included); even if the author seems credible.

Here is Intel's official statement on min/max voltages:"Table 2-2 specifies absolute maximum and minimum ratings only and lie outside thefunctional limits of the processor. Within functional operation limits, functionality andlong-term reliability can be expected. At conditions outside functional operation condition limits, but within absolutemaximum and minimum ratings, neither functionality nor long-term reliability can beexpected. If a device is returned to conditions within functional operation limits afterhaving been subjected to conditions outside these limits, but within the absolutemaximum and minimum ratings, the device may be functional, but with its lifetimedegraded depending on exposure to conditions exceeding the functional operationcondition limits.At conditions exceeding absolute maximum and minimum ratings, neither functionalitynor long-term reliability can be expected. Moreover, if a device is subjected to theseconditions for any length of time then, when returned to conditions within thefunctional operating"This is the other unfortunate side of buying a QX processor. There doesn't appear to be any more stress tolerance built in, though clearly it is marketed to folks who will push it significantly as they rated QX processors w/ a Thermal Design Power of 130W versus 95W for the straight Q processors. They're pretty clear to warn you if you run it outside of the 'functional operation limits', you are taking on some known risk.