Faulty temp reading?

[Angilion]

I'm running a new X2 3800 at stock, low case airflow, EVO 120 cooler, MX-1 paste. I had trouble aligning the cooler and I think that's messed up the paste, causing temperature problems. Or maybe not...

I tested using S&M maximum CPU heating test, with all tests selected on both cores. That reports an almost constant 100% load on both cores.

Speedfan 4.32 is reported odd core temps, varying up to 14C in a matter of seconds. I think that "temp 2" is one core and "core" is the other.

S&M's monitoring shows only one temperature reading from the CPU, roughly matching temp 2 reading from Speedfan.

I'm sure that the temperature of the cores, especially the second core, can't be varying that much that quickly. A glitch in Speedfan?

 

[Craig321]

Give Everest home a go to compare it's temperature readings to Speedfan: http://www.majorgeeks.com/download4181.html

 

[Angilion]

I did. It doesn't fully support my unusual motherboard and only shows one temperature for the CPU.

 

[Angilion]

The plot thickens...

Immediately after 20 mins of S&M, with Everest and S&M reporting a CPU temp of 53C and Speedfan reporting 53C and 49-62C, I rebooted into the BIOS. Hardware monitoring in the BIOS reported a CPU temp of 43C.

I overclocked to 2.5GHz. Straight up from 2.0 to 2.5. Booted fine, ran S&M again. Temps went to 56C in Everest and S&M, 56C and 50-66C in Speedfan.
Stopped S&M, rebooted immediately, back into BIOS...hardware monitoring in the BIOS reported a CPU temp of 44C.

Frankly, I have no idea what my temps are. That bothers me. I suspect that I misapplied the thermal paste because I have trouble aligning the cooler and I think doing so smeared the paste all over the place...but maybe it's fine and the temps are about 45-46C under full load (allowing 2C cooling in the 30s or so before shutting down and reaching the BIOS).

That "Core" reading in Speedfan jumps 12C as soon as S&M is started. That can't be right.