A REPRISE: The Wolfdale temp-sensor problem

BonzaiDuck

Lifer
Jun 30, 2004
16,830
2,147
126
Just when we had a cool respite in So-Cal, we're back in a doldrum of 95F afternoon temperatures. My digital thermometer in the "computer center" climbed to 84F -- calibrated, that means 80.9F.

Added back into my system a second gfx card for SLI, and wanted to stress-test-certify again (although the prevailing advice might suggest that -- in theory -- it's unnecessary.)

The "stuck" Tj core temperatures were rising to 71C/71C. No problem there -- if you can trust those values -- since they're stuck at 56/56 in idle.

But this time, I loaded Everest Ultimate 4.50 while running INtelBurnTest. I notice that the TCASE temperature is EQUAL TO the core temperatures. This ain't the way it's supposed to be. (And you would wish that RealTemp would report TCASE as well as the core Tj's.)

There are various possible scenarios. Among them, the latest BIOS version is not accurately reporting TCASE, even if the variation in TCASE is what one would expect.

Of course, there's another scenario that I've come dangerously close to the CPU thermal spec.

But the latter scenario doesn't make sense -- if I can believe BIOS monitor, Everest, CPU-Z, etc. for showing a load CPU voltage of 1.24V. This system is only over-clocked by 20% -- from the E8600 default of 3.33 Ghz to 4.0 Ghz. None of the other voltages are "out-of-whack," either.

Caller! You're on. Comments?
 

Dadofamunky

Platinum Member
Jan 4, 2005
2,184
0
0
Not sure it's a problem, wasn't the Wolfdale supposed to fix the sensor issues? My settings are very close to yours and I'm pretty disappointed with my temps. I can run OCCT stress testing but I'm averaging 63C at load. 42-43 at idle. Not very happy with that. Might just have to live with it.

Hey SoCal, stop stealing our water! :D

(I'm from NoCal)