What's new
  • Please do not post any links until you have 3 posts as they will automatically be rejected to prevent SPAM. Many words are also blocked due to being used in SPAM Messages. Thanks!

CPU temps vs Core temps

sloth

Member
Joined
Oct 16, 2008
Messages
9
Hi, I just got my first rig up and running. It is all running pretty smoothly except I have some questions about CPU temps.

I'm currently running an e8400 with the stock heatsink and for now the side panel of my silentium T3 case is off. The readings are as follows when idling (all in celsius):

System 29
CPU 35
AUX 29
HD 45 (My case has a weird muffler system which essentially encases it in foam with a heatsink on the side, I think I might do away with it and put it in a regular bay)
Core 0 51
Core 1 52

Are these normal? The thing that concerns me most is the flame icon next to the core temps (flames are bad for computers right?).

Other things that bother me are that once, when I left for a couple hours when formatting the drive for the first time, I came back and it said there was an error saying the processor overheated, press f1 to continue.

Also, when I had some colour calibration process running which made both precessors go at 50%, the CPU temp went to 88. (I uninstalled it and now I have these more normal readings, but it concerns me that a 50% load would send my CPU temp to 88.


Any thoughts would be much appreciated, I'm kind of new to DIY computing?

Thanks
 

Jon_di2

Well-known member
Joined
Jul 16, 2008
Messages
876
Location
Toronto, Ontario
Core 0 51
Core 1 52

Are these normal? The thing that concerns me most is the flame icon next to the core temps (flames are bad for computers right?).
yes flames are bad... Normal to see the flame under load, but you should never see it while idle. Something has gone wrong here... Even a stock heatsink should not be running temps that high. These temps are not normal! Did you mount it yourself? If so, did you follow a guide? Personally I wouldn't use a stock heatsink anyways, you should get a good heatsink and re-do it.

reasons I would think your getting high temps:

1) Your using pushpins with your heatsink and they are not pushed in all the way
2) the TIM was applied improperly... or not at all....
3) The fan isn't running...
4) bad overclocking...?

ps: stop using the computer if your running temps under load over 70!! 88 is not good at all!!!!
 

sloth

Member
Joined
Oct 16, 2008
Messages
9
Thanks for the reply

Im not overclocking so that could not be it.

I have tried to reseat the heatsink but now the temps are even higher.


I fell like when I reseat the heatsink I am messing up the thermal paste (pushpins are the devil). Should I be reapplying the thermal paste every time I reseat the heatsink? Right now there seems to be a solid connection but the temps are higher than beofre.

cpu 41
Core 0 57
core 1 58

All of these readings were with speedfan so downloaded Real temp and it gives me core temps about 10 degrees lower (but no CPU reading). Maybe I shouldn't have reseated it in the first place.
 

enaberif

Well-known member
Joined
Dec 9, 2006
Messages
11,391
Location
Calgahree, AB
Try latest CoreTemp or RealTemp.
Idle temperatures for 45 nm CPU's are way off. All that matter is distance to TJMax in full-load (LinX/IBT or Orthos). Keep it >20-30C and you're fine.
They aren't that bad :blarg:

Idle temps are very dependent on ambient temp.

For dual core cpus a idle temp between 30-40 celsius is perfectly acceptable and a load temp between 60-70 is fine as well.

Use P95 or Orthos with Small FFT test.
 

sswilson

Moderator
Staff member
Joined
Dec 9, 2006
Messages
17,897
Location
Moncton NB
They aren't that bad :blarg:

Idle temps are very dependent on ambient temp.

For dual core cpus a idle temp between 30-40 celsius is perfectly acceptable and a load temp between 60-70 is fine as well.

Use P95 or Orthos with Small FFT test.
88c under load isn't all that bad???

I wouldn't expect to see anything over 55c on a non-overclocked C2D processor. Something is definately going on, and if you can't find it, it's time to find somebody who can.........

As to the original question... CPU temp means very little when it comes to a multi-core cpu. The # everybody pays attention to is the core temp.
 

burebista

Well-known member
Joined
Sep 14, 2007
Messages
594
Location
Romania
They aren't that bad :blarg:
Man, believe me, I spoke alot with unclewebb (RealTemp developer) and from his experience and others it looks like on 45 nm DTS readings are good above 60°C and almost very good near to TJMax when thermal throttling should kick in. In idle they are way off.
unclewebb said:
These sensors weren't designed to accurately report core temperatures. Intel designed them to trigger thermal throttling and thermal shut down. There is a small amount of error where Intel calibrates them near TjMax and they can be off by 10C or more at idle even when they aren't sticking.
And that's why I like RealTemp over CoreTemp
unclewebb said:
A lot of users think that using the correct TjMax is automatically going to give them accurate temperatures but that idea is wrong. There is so much error in most 45nm sensors, both at idle and even at the calibration point near TjMax, that you need to do some sort of user calibration to have any hope of reasonably accurate temperatures and unfortunately, that's the best you can hope for.

Intel hasn't released enough information about these sensors for any software developer to translate the data coming from the on chip sensors into 100% accurate core temperatures. The 45nm sensors simply aren't good enough for that. Having said that, I still think that a RealTemp type calibration is better than nothing and will result in some reasonably accurate reported core temperatures.
I have my example a couple of minutes ago. 23.75°C room temperature.

RT_compare1.png

So anyone believe that an E8400 C0 in idle at 1586 MHz and 0.944V Vcore (5W power draw) cooled by Ninja fanless+ TR Bolt Thru and kept in an Antec P182 with Noctua exhaust and Scythe Kama PWM intake can have 40°C on cores?
For me the only thing that matter is distance to TJMax. As you see in both cases is the same 58, so I try to keep it >20-30 in full-load and nothing else matter.
 
Last edited:

sloth

Member
Joined
Oct 16, 2008
Messages
9
Thanks for all of this so far, you guys are a great help.

Ok, SO I reapplied the thermal paste (arcticsilver 5) and it now idles at 39 with distance from tjmax at around 58. All seems well when it idles.

I ran IBT and I could not run a 1/2 stress test more than once without crashing. I tried orthos first and it said fatal error and wouldn't finish the test. I then tried to run a full stress test with IBT and I got a blue screen that told me about a fatal error and dumpoing the physical memory.

This is all with the default settings (no overclocking)

It might be worth mentioning at this point that I am running Win XP pro 64-bit (I want to realize all 4 gigs of ram and I don't like vista).

Anyway... how bad is this? I'm not really a power user, I play the occasional RTS and run MATLAB sometimes. I just like to know that everything is ok especially since I put this rig together and I want to know if all the CPU fine.

Finally, is all this stress testing hurting my processor if it is infact overheating?

Should I be contacting NCIX (where I bought it) about it?

Thanks
 

Latest posts

Twitter

Top