Real Temp - New temp program for Intel Core processors

[unclewebb]

AdmiralThrawn: Sounds like you were getting close to throttling but a properly functioning E8400 will not crash at 95C at default MHz and voltage. There is still lots of head room left before that happens. If you are overclocking and overvolting and running something like Orthos then you'll have an impossible time getting up to these temps without a crash but when lightly loaded, it's a different story.

Thermal throttling seems to start to happen when DTS drops down to 2 or 3. That's when the PROCHOT# signal will be activated which was reported correctly by RealTemp during my testing.

Vinas: Everything I believed about the relationship between Intel specified thermal limits and actual TjMax were thrown out the window yesterday when I was testing a revision M0 E2160.

Intel Thermal Specification
TCase Max:
E2160 73.2°C
E8400 72.4°C

Going by those specifications a person would be led to believe that TjMax for the E2160 is at least as high as the E8400 and maybe even a degree or two higher. That is not the case.

The E8400 has a measured TjMax=95°C but the E2160 during the same kind of testing is only TjMax=85°C.

Lesson learned is that you can't look at the Intel Thermal Specifications and determine anything about TjMax.

RealTemp v2.2 should be available later tonight which will include a fix for the M0 processors.

I now have serious doubts about the E6x50 series but until I actually test one I'll be leaving them at TjMax=95°C. RealTemp v2.2 will allow setting a TjMax offset of +/- 5, 10, 15 for those that wish to experiment or don't believe the RealTemp chosen TjMax.