Sign in with
Sign up | Sign in
Your question

CPU temp vs CORE temp (post your temps!)

Last response: in CPUs
Share
a b à CPUs
February 26, 2010 9:59:26 PM

I've replaced my old Athlon X2 5200+ with an Athlon II X3 435.
By performing the upgrade, I got to know the CPU temperature sensor in my motherboard (Asus M3A) is broken (it reports higher temps than the actual temps), since I installed my old X2 in a friend's computer and it was like 45º C under load instead of 65º C in my mobo.

So, the Core Temp sensors in the II X3 seem OK, since they get up to 43º C under load at most (CPU Temp is 66º C under load).

I would like to know what my CPU Temp should roughly be according to the Core Temp sensors. Hence, I ask you which are your corresponding CPU and Core temps for your processors (at load, preferably)? I'm afraid the comparison may be only valid with AMD's Phenom II and Athlon II CPUs, but I invite anyone to post their temps.

More about : cpu temp core temp post temps

a b à CPUs
February 26, 2010 10:05:24 PM

E1400 OC'd to 2.4ghz @55c load.
m
0
l
a b à CPUs
February 26, 2010 10:50:03 PM

Q9550 @ 3.8 1.26v loaded is 48-52c (on air) on the cores.

Will have to look up the cpu temp when I get home (at work on my droid, as I am whenever I do toms lol, usually), but from what I remember its usually close to my core temps... I think.
m
0
l
Related resources
a b à CPUs
February 27, 2010 12:43:33 AM

Thanks for the feedback. Anyone with an AMD cpu???
m
0
l
a b à CPUs
February 27, 2010 1:19:49 AM

Let's see, my Athlon II X2 250

Core Temps: 13C (55F) which is below ambient and wrong.

CPU Temp: 25C (77F) which is about 12F above ambient and probably correct.
m
0
l
a b à CPUs
February 27, 2010 1:23:09 AM

Those are at load??
m
0
l
a b à CPUs
February 27, 2010 1:35:45 AM

Sorry, no. I'll try that now for ya.
m
0
l
a b à CPUs
February 27, 2010 1:39:50 AM

At load:

about 40C on CPU and 29C on cores. I believe the CPU temp to be more correct.
m
0
l
March 8, 2010 7:37:33 PM

I haven't looked into it too much, but I believe there are sensors on each core and one behind the CPU socket. The core ones are part of the CPU while the socket sensor is part of the motherboard.

Anyway my PhII 925 stock @ load: Socket: 39C Ambient: 33C Cores: 33C
m
0
l
a b à CPUs
March 8, 2010 8:15:10 PM

The PII may just have one core sensor (rather than 4).
m
0
l
March 8, 2010 10:19:51 PM

EXT64 said:
The PII may just have one core sensor (rather than 4).


I doubt that. Even my old Athlon 64 x2 had individual core sensors.
m
0
l
a b à CPUs
March 8, 2010 10:33:08 PM

Yep, the Athlons had individual (2), but the PIIs do not (only 1).

"**CPU Temp is because the Phenom\Opteron (K10) have only one sensor per package, meaning there is only one reading per processor. "

http://www.alcpu.com/CoreTemp/howitworks.html
m
0
l
March 9, 2010 3:23:09 AM

So why does my PhII show 4 individual core, that occasionally have different values in HWomnitor?
m
0
l
a b à CPUs
March 9, 2010 7:51:23 PM

Yep, they show up as 2/3/4 sensors but they all read the same on Athlon IIs and Phenom IIs.
m
0
l
March 9, 2010 8:21:10 PM

Athlon II X4 620 42 degrees idle.
m
0
l
March 9, 2010 9:55:09 PM

sanchz said:
Yep, they show up as 2/3/4 sensors but they all read the same on Athlon IIs and Phenom IIs.


Mine will vary from each other occasionally. This isn't the greatest proof but as you can see in my photo on the minimum temp in HWmonitor 3 of the cores dropped down colder than the other.

m
0
l
a b à CPUs
March 9, 2010 10:19:54 PM

Hmm, that is interesting, however not nearly as different as it should be. On CPUs with 4 different sensors, it is extremely rare to have all 4 cores the exact same temp. I have seen them different, even on my PII 940. However, it is never more than one read cycle. I wonder if it may be due to the program refreshing (taking a new sample) at the same time the sensor is updating. So rather than recording temperatures in different places, it is reporting temperatures at different times.
m
0
l
March 9, 2010 11:31:41 PM

EXT64 said:
Hmm, that is interesting, however not nearly as different as it should be. On CPUs with 4 different sensors, it is extremely rare to have all 4 cores the exact same temp. I have seen them different, even on my PII 940. However, it is never more than one read cycle. I wonder if it may be due to the program refreshing (taking a new sample) at the same time the sensor is updating. So rather than recording temperatures in different places, it is reporting temperatures at different times.


That sounds reasonable. My Athlon 64 x2 5600+ cores rarely varied, but on my brothers Athlon 64 x2 5050e I noticed about a 5C degree difference between the cores on load.
m
0
l
a b à CPUs
March 10, 2010 12:17:59 AM

Athlon 64s had 2 CPU temp sensors...
m
0
l
March 10, 2010 12:40:37 AM

Yeah. I caught that the first time it was said. So how come AMD cheaped out and only put in 1?
m
0
l
a b à CPUs
March 10, 2010 12:47:29 AM

BradyT88 said:
Yeah. I caught that the first time it was said. So how come AMD cheaped out and only put in 1?


Cuz it's cheaper. And people rarely use the cpu-temp sensor anyways. It's far more practical to look at the core temps.
m
0
l
March 10, 2010 12:47:48 AM

BradyT88 said:
Yeah. I caught that the first time it was said. So how come AMD cheaped out and only put in 1?


I'm guessing heat conducts so quickly between the cores since their on the same die that they can pretty much all be considered the same temperature.

But anyways my 955 @ 3.9 is 26 idle (with cool and quiet) and 55 load. Supposedly anything under 61 is good, but stability starts to fail around 52. I don't even use cpu temp from mobo.
m
0
l
a b à CPUs
March 10, 2010 1:34:46 AM

I have no idea why they only put one in. I can't imagine it was that expensive, and I like more sensors as it can tell you if your thermal paste is applied properly and evenly. The temp variation should be small (only a few C) but it is still nice to see. My FX-60 can vary quite a bit at times, up to about 12F.
m
0
l
March 10, 2010 2:13:52 AM

XD_dued said:
But anyways my 955 @ 3.9 is 26 idle (with cool and quiet) and 55 load. Supposedly anything under 61 is good, but stability starts to fail around 52. I don't even use cpu temp from mobo.


What is your vcore set at and what kind of cooler do you have? My 925 at 3.8GHz @ 1.55v has a corsair H50 on it and temps are about 33/49 idle/load.
m
0
l
a b à CPUs
March 10, 2010 11:14:43 AM

So, does CnQ still work when OC'ed? I know my 940 could not use CnQ when undervolted (as it is now to 1.25V (from 1.35V)).
m
0
l
a b à CPUs
March 10, 2010 11:41:41 AM

BradyT88 said:
Mine will vary from each other occasionally. This isn't the greatest proof but as you can see in my photo on the minimum temp in HWmonitor 3 of the cores dropped down colder than the other.

]http://img384.imageshack.us/img384/7099/sensors.th.png

HWMonitor is messed up. I don't know what the developers are trying to achieve but the readings I've seen on some Core i7 processors are unreasonable (and no application should, out of the box, report different temps to the rest for i7/i5).

Oh, and yes, Phenom II has only one DTS per package.
m
0
l
!