What is the difference between core temperature and CPU temperature?

tsd16

Distinguished
Jan 11, 2008
333
0
18,790
As the title says, where does the CPU temp come from?

i.e.

I have four cores ranging between 30-35C idle while the cpu temp reads 44C, on everest for example. Why would the CPU temp read higher then the cores, which are on average 10C cooler?

I have used about 4 different programs core temp real temp everest speed fan. some only show the cpu temp some only show core temp. everest was the only one to show cpu temp and core temp.
So on everest I get a read out of

CPU Temp: 44C
Core 0: 32C
Core 1: 35C
Core 2: 30C
Core 3: 33C
 

tsd16

Distinguished
Jan 11, 2008
333
0
18,790


Thanks, by some of the literature there, my temp readings seem to be contradicting what it says:

Heat originates within the Cores, where Tjunction sensors are located on the hot spots of each Core. Most of the heat dissipates from the top of the Cores through the Integrated Heat Spreader and CPU cooler to air inside the computer. Some of the heat dissipates from the bottom of the Cores through the CPU case, which creates a 5c thermal Gradient toward the center of the substrate, where the Tcase sensor is located. This heat then dissipates through the socket and motherboard to air inside the computer. Safe and sustainable temperatures are determined by CPU cooling efficiency, computer case cooling efficiency, Ambient temperature, Vcore, clock speed and Load.

* Tjunction is higher than Tcase.


This is not the case with my temp readings, Tjunction is lower than Tcase.

Although I am running an i7 processor I don't know if this data would be different in this case.

From what I gather, after googling" Tcase higher than tjunction", some motherboards read Tcase inaccurately. I still don't like it.
 

Zorg

Splendid
May 31, 2004
6,732
0
25,790
I believe that Computronix is working on an i7 update or guide. The Tjunction should be higher than the Tcase.

Try checking the idle Tcase in BIOS and compare it to the one you are getting from the program, they should match reasonably closely. It is worth a look to be sure you are actually reading the Tcase sensor. You can also run Prime95 small FFT torture test for a few minutes and track the Tcase to see that it increases.

The core Temp monitoring programs could easily be using a Tjmax that is incorrect. There could be sensor anomalies as well.

I would keep the Tcase under 70C, after you have confirmed it is the right sensor, until you get it sorted out.

 

tsd16

Distinguished
Jan 11, 2008
333
0
18,790



Tcase temp does match whats read in the bios, so every piece of software I am using and the bios support eachother, im not getting differing values, so my board appears to be actually reading it at 10c higher.

I ran a 10 min prime 95 test at max load tcase didnt break 65c, and the cores didnt break 59c.


Like I said i searched the net and have seen (primarily with asus boards oddly enough) many cases where the tcase is being in accurately read i.e. a few degrees higher than the cores, instead of lower.
 

tsd16

Distinguished
Jan 11, 2008
333
0
18,790
Well I updated the bios to the latest version and there is for sure a change

Tcase idling at 37c

Cores same 30-35c.

So its still higher, but that at least tells me something is wonky with the sensor or the way the mb interprets it. (yeah i made sure to turn speedstep off, i forgot at first and cores were in the 20's and the tcase was in the low 30's.)

Under load, the Tcase never gets hotter than the hottest core 59c. so thats a 6c difference under load and about 7c idle. although it is night time and my pc room is a bit cooler. But my system (board temp) is hovering around 42/43c what it always seems to be at idle no matter what the ambient is.

I mean since Tcase "should" be cooler than tjunction is there anything you can think of that would cause that temp to be higher? Other than kafluky temp readings?
 

Zorg

Splendid
May 31, 2004
6,732
0
25,790
No, it is physically impossible. I would imagine that the process of calibrating Tcase to ambient (room not mobo temp) and then Tjunction to Tcase will be similar. That is the only way to ensure that the temps are accurate. At any rate your temps look good to me, so I would wait for the guide and or another BIOS update. I would imagine that there is data on the web about this by now. You might also try the Real Temp forums.
 

Zorg

Splendid
May 31, 2004
6,732
0
25,790
As I said in your other thread, the above guide has been updated so you can calibrate your sensors. You will have to wait until Tom's gets the full guide posted.
 

tsd16

Distinguished
Jan 11, 2008
333
0
18,790



I assume you are reccomending there to power the system down for awhile (long enough to come down to room temp) power it on and immidiately check the tbase in the bios. and guestimate that the CPU should be at least within a couple degrees of room temp actual.
 

tsd16

Distinguished
Jan 11, 2008
333
0
18,790


ah, k thanks.
 

Zorg

Splendid
May 31, 2004
6,732
0
25,790
No, not powering it down. Just use a thermometer to make sure the room temp is correct.

I saw in your other post that you were going to try to calibrate your sensors. You need to wait until the whole guide is back up and read the calibration section and follow it exactly.