-X-hellfire: In case you don't know, a Core 2 Quad consists of two separate Core 2 Dual cores. Your data looks very similar to my Q6600 where I am about 99% certain that TJMax on core 2 and core 3 is set 5C higher than core 0 and core 1. I have seen too many similar screen shots to believe that this is all just random chance. I don't believe that Intel has told users the whole story about how they set TJMax on their 65nm Quads.
For your CPU I'd try TJMax 100, 100, 105, 105 and then use slightly different calibration factors on core 2 and core 3 to balance out the temps at idle. I think after this, your 4 cores should track each other very closely from idle to TJMax when running Prime 95 Small FFTs. I like that program and test because of its consistency. My Quad still shows this 5C difference between the two sets of cores even when the temperature is at 90C or 95C. That doesn't seem like slope error. That seems like TJMax is being set differently by Intel. Once calibrated, you can do the low MHz / low voltage test and compare your reported temps to your room / water temp.
Snafu: Talking temps is never off topic on this thread. It hasn't always been easy turning a couple of lines of code into a 127 page story.![]()
Bookmarks