Q9450 temperature variations

M2P

Distinguished
Jul 22, 2010
11
0
18,510
Hi people,

I have recently installed a q9450 and it has a variation of 15 degrees at idle/general use i.e. 35/50/50/35
This does not vary in either occt/core temp/ real temp.
When stress testing with prime or occt, the variation narrows to 10 degrees. (celcius)

My other recent 775 chips and my i7 920 never really vary more than 5/6 degrees at worst.

Should I be concerned?

Thx
 
Solution
tj max may be around 100'C, but CPUs will still degrade and be damaged in the 90s'C or upper 80s'C.

To the OP:
You probably have a stuck temperature sensor. As long as you're under 80'C for stock or low 80s'C for overclocked benchmarking, you should be fine.

jeroly

Distinguished
Jun 27, 2007
113
0
18,710
My understanding is that the 100C threshold is being measured differently (at a different place in the chip) than the readings you get through monitoring programs like SpeedFan, and that you really need to keep it under 85C or so.

I'm running an overclocked Q6600 (@3.2Ghz) which has the same 15C variation; it's about 43C idle, 58C under load. No problems.
 
tj max may be around 100'C, but CPUs will still degrade and be damaged in the 90s'C or upper 80s'C.

To the OP:
You probably have a stuck temperature sensor. As long as you're under 80'C for stock or low 80s'C for overclocked benchmarking, you should be fine.
 
Solution

M2P

Distinguished
Jul 22, 2010
11
0
18,510
Ok guys, thx.

The variation doesnt really alter whether under load or at idle.
Bluescreendeath, i think you have got it.
Do you think it's worth rma'ing now as it only has a year left of intels 3 year warranty or do they not cover this sort of problem/variation?
Overclocked, it rarely cracks 63 degrees under stress tests and I have ran prime95 for up to 6 hours.
 

Nils

Distinguished
Aug 7, 2008
586
0
19,010

Those stress temps are really fine. You should have nothing to worry about. Maybe the core temp sensors have different calibrations.