I7 2700K Sandy Bridge core temps false?

Welsher

Honorable
Jan 18, 2013
16
0
10,510
Hey all, i'm new here so please bear with me and the gigantic wall of text. I apologise if this is in the wrong part of the forum.

So recently I built a new computer (will post specs below) and for days I enjoyed playing the likes of BF3 on high settings for hours on end with no problems.

Out of nothing but curiosity and boredom (I doubt I would have ever discovered this problem otherwise), I soon decided to check the temperatures whilst in game. I found that my core temperatures would often reach the TJmax for my particular CPU (in this case it is 98c) within a matter of 2-3 minutes. I used several programs to monitor my temps including Core temp, Real temp and HWmonitor in order to confirm this - they all report the same. I thought originally that these high temps were caused by high Vcore voltage (it was 1.44v at the time), however I have since replaced the PSU (which fixed the overvolting problem) but I still have high temps. Before replacing the PSU, the BIOS would tell me that the CPU temperature was 42-45c - however it is still the same even with my new PSU. I have checked everything else, I have made sure the thermal paste is applied correctly, that the plastic cover for the CPU was removed, that the cooler is on correctly etc.

After trying all this, it panicked me but I realised that had these readings been accurate, surely I would have received a warning or notification of some sort or a symptom from reaching TJ max (such as CPU throttling)? Yet I have not had any problems whatsoever - not a single hiccup; it has been working flawlessly.


My actual CPU temperature appears to be well within Intel's thermal specifications for this chip when under load and when idle (under load I'm in the 50s, idle in the 30s - Tcase for this CPU is 72c). This is a screenshot of HWmonitor that I recently took whilst running a standard Intel Burn test;

UUrP0.png


So all in all... do you think this is a genuine problem, or a false positive?

PC specs:

I7 2700K Sandy bridge (not overclocked)
GTX 460
Seasonic 550w modular PSU
MSI Z77A G43



Thanks in advance.

 

kenny3105

Honorable
Oct 20, 2012
289
0
10,860
are you using stock heatsink if so make sure the plastic pins have been pushed in till you hear a click also are you using cpu gfx if so disable it in bios and the temps will drop
 
Softwares to monitor your temperatures are always a Little unacurrate, BIOS is the most reliable. Though when your temps are so high, I believe you have a problem which needs to be solved quickly.

It could be either your airflow in the case, the ambient temperatures or the fan itself, that could cause this.
 

Welsher

Honorable
Jan 18, 2013
16
0
10,510
That program you suggested just listed 2 of my cores as throttling, and my actual CPU temperature as 97c after just seconds of the standard Intel burn test.


Yes I am using the stock heatsink, the pins are all secured - pretty sure im not using CPU gfx, will have to check


@lostgamer_03

Intel support told me the BIOS reading was unreliable - regardless it reads 42-45c constantly. Atm the side of the case is off, the ambient is below room temperature - I'm going to get a better CPU cooler soon, see if that fixes my problems. Although the stock fan should be sufficient.
 


BIOS are not unreliable. You can't see the temperatures under load though that might be what Intel support meant. What about the thermal paste? If you have used too much, then the heat can't get away, because it's getting isolated. :)
 

Welsher

Honorable
Jan 18, 2013
16
0
10,510
I should of copied the conversation, I've completely forgotten why he said it was unreliable - I was taken aback by it at first because in the other forum I posted this issue in, everyone told me the BIOS was reliable xD

The stock intel heatsink comes with thermal paste pre-applied
 


Oh I forgot! Some MOBOs doesn't increase your CPU fan when it gets hotter, so that might be where the problem is coming from. Your fan might run at only one setting.

Try to boot into BIOS and check or look it up in your manual. It might save you a lot of time instead being in need of replacing the CPU and likely see the same problem again.
 

Welsher

Honorable
Jan 18, 2013
16
0
10,510
Well I look like a complete spastic, one of the pins on the cooler wasn't in properly - I was afraid to push it further because it felt like it would break the mobo - I noticed one pin wobbled slightly whereas the others didn't so I pushed it more and it clicked. :lol:

I now idle in the late 30s to early 40s, I completed the intel burn standard stress test with a max CPU temp of 72, and max core temps of 70. Whilst that is a massive reduction, it is still on the threshold of what is acceptable under load - could this be due to the thermal paste messing up due to the cooler not being on properly? I might still end up buying a better cooler and thermal paste. I played a bit of BF3 too and my temps are in the 50s/60s now, all seems to be good. Thanks for the suggestions and replies anyway guys, appreciate it
 


Silly, silly, silly.. ;)