Sign in with
Sign up | Sign in
Your question

CPU core temp disparity

Last response: in CPUs
August 17, 2008 10:47:27 AM

I just built a comp with a q9450 on a gigabyte x48-dq6 with a Big typhoon HS
Using CoreTemp and RealTemp, one of my core's temps is higher than the other cores' temps.
On idle 3 of the cores run at 28C whilst the 4th core runs at 52C!!
I also noticed the 4th core's temp never changes even on load.
I suspect a sensor is stuck.
Has any1 experienced this before and will it prevent me from overclocking?

a b à CPUs
August 17, 2008 11:32:26 AM

Has to be a stuck sensor
August 17, 2008 12:00:27 PM

Real temp has a "Test Sensors" Button bottom left. When you say the 4th core sensor doesn't move have you had the cpu running at 100% for an extended period, you should be able to get it past 60c and if it doesn't move you have an issue and should return it. This issue doesn't effect overclocking as your concerned with max temps.
There are some hit and miss ways of unstucking sensors which invovle max loading for hours and sometimes days.
Related resources
August 17, 2008 12:56:44 PM

Thanks, I will run Prime95 for a few hrs and see if that changes anything.
What i find unusual is that it is 20C higher than the others, could it be that i havent seated my HS properly?
I guess it could be the cooler IF the 4th core temp goes over 52C after its been on load for a few hrs.
Does a stuck sensor warrant a replacement?
August 17, 2008 1:13:13 PM

Generally if a sensor is stuck on the way up you have a case, but Intel doesn't warrant sensors that are stuck at idle as far as I know. It's unlikely that a 20c difference is caused by an unseated HSF but it could be possible, what you need to do is gently "rock" the HSF at all four feet because more than likely 3 have seated properly. Good luck, hope it does become unstuck, can you please report back if it does become unstuck, as I am trying to roughly gauge what success rate this method has.
August 17, 2008 1:26:36 PM

or maybe because the the programs don't read the temperature correctly because of the DTS specification in which Intel is releasing this week at the IDF.
August 17, 2008 1:33:57 PM

Okay, im running Prime95 atm and will report back later.
August 17, 2008 2:41:51 PM

so far the temps are:
core #0: 31C, core #1: 31C, core #2: 52C (the one that is stuck), core #3: 40C

It will take some time to get my cpu hot enough. i have turned down my HS fan speed.

What i find interesting is that core #3 is 9C hotter than the #1 & #2, this could mean i have properly seated my heat sink.
August 17, 2008 3:02:28 PM

I think your going to have to think of something else if your cores are only going from 28 to 31c. Your going to have to get the cpu past 60c for an extended period just to have a chance at unstucking it. You might have to turn all the other fans off or even go back to the stock cooler with the fan at low.
August 18, 2008 1:33:54 AM

yeah i know, this just becoming stupid. in my last post i meant *havent properly seated my HS, so i will re-install the stock HS. Cant wait for intel to release DTS specs.