TR Forums

Hi Guys, I have a question about temperature monitoring software. Until a while ago i was using Speedfan 4.37 and i was quite happy with it. It used to monitor my temps and fan speeds and with the latest version it also supported GPU monitoring. Based on its reading's, i even changed my q6600 heatsink (used to show my quad core at 69-73 C at full load at 3.2GHz) and everything was fine. Recently, a friend of mine suggested using Everest to monitor my temps since it also monitors the hard disk temps. Thats when it started to go wrong. While Speedfan said my temps were around 34/57 at idle/Load, Everest lists them as 25/46. Thats a huge difference. So i tested both on my brothers Gigabyte motherboard and the same thing happened (though not the same exact delta temp's, but that could probably be because he uses a different CPU). Funny thing is, Easy Tune (Accesory bundled with Gigabyte motherboard) showed different numbers than the other two software( though its numbers were very close to the everest numbers). My question is a.) which of the software is reporting the right temps and b.) if none of them are right, could you guys recommend a software which is. I'm really confused.

Coretemp comes to the rescue

Oh and welcome to The Tech Report

edit:

Although I should say that software is just for the CPU and nothing else, GPU-Z does do temp monitoring for GPU's if I'm not mistaken hehe

Try CoreTemp and RealTemp. Everest is generally very good so compare results. Take the average and see which program comes closest.

Thanks a lot for the response guys. I just used Core temp and used Prt Scrn to compare the results at the same instance. Here they are

Speedfan: 51/52/49/43.

Core temp: 51/50/47/43.

Everest: 44/40/36/36.

Based on the above results, is it ok to assume that Everest is wrong?

And why the differences in temps. Don't they all use the same sensors? And if Everest is in fact wrong, can i trust the hard drives temps it reports ( For what it's worth both Everest and Speedfan display my GPU temp as 56 c)? Thanks again.

The temperature sensor on Core2s does not report absolute temps, just degrees celcius relative to the chip's max operable temperature (or TjMax). Intel hasn't officially published the TjMax values for your Q6600 chip so it's left up to educated guesses and speculation on the part of the software author to determine the correct(ish) value. I usually use Core Temp or Real Temp for cpu thermal monitoring.

Everest seems pretty accurate for me.

RealTemp:

63 / 64 / 62 / 57

CoreTemp:

62 / 63 / 61 / 58

SpeedFan:

47 / 47 / 46 / ?

Everest:

62 / 61 / 61 / 57

Core i7 3.6Ghz during Intel Burn Test.

SpeedFan seems wonky...

mjallan123 wrote:

Everest seems pretty accurate for me.

It may depend on which version of Everest we are talking about, because they keep changing the guess for the various Tjmax'es.

@crazybus, I thought they did release Tjmax for C2Q's?

Flying Fox wrote:

@crazybus, I thought they did release Tjmax for C2Q's?

Only for 45nm Core2s IIRC.

http://www.overclock.net/intel-cpus/375827-attention-all-intel-s-official-tjmax.html wrote:

TJ MAX SPECS:

Intel®Core™2 Duo processor E8000 and E7000 series: 100°C
Intel®Core™2 Quad processor Q9000 and Q8000 series: 100°C
Intel®Core™2 Extreme processor QX9650: 95°C
Intel®Core™2 Extreme processor QX9770: 85°C

Pertinent Intel slide available here.
Of particular note is that the DTS is calibrated individually per chip so there is sample variation. Idle temps are also inaccurate due to slope error at lower temps.