Q9650 core temp differences? How much is normal?

[GRAFiZ]

Got my Q9650 setup the other day, Gigabyte P45, 2x2gb $25 Patriot DDR2!!!

Anyways, seems to be fairly happy at 4ghz which is all I've tried for, that was my original goal.

It's watercooled, on a single 1x120 rad.

The temps seem good under load, as well as idle, but the differences in the cores are rather extreme.

I've pulled the block off several times, applied AS5 using the single line, double line, single pea and double pea method... different amounts of tension etc.

Heres a picture of the best I can get it. Should there be such a large swing in temps between the cores or is something wrong? Also, whats the max temp I should allow this chip to reach? 65c?

Thanks all!

 

[unclewebb]

Large differences between individual cores on a 45nm Core 2 Quad are normal. Part of the difference is slope error where the sensors move at a slightly different rate than the core temperature changes at and the rest of the difference is differences in TJMax from one core to the next. Intel has admitted that TJMax is not a fixed value but didn't publicly state how big this range can be. In my testing, a 10C difference between two cores on the same CPU is not out of the question. From what I can see, your sensors look far better than most. Post that CPU Cool Down Test. It can help spot things like sticking sensors.