Would my true CPU temp please stand up?

[kumquatq3]

I have a lapped q6600 with a Xigmatek s1283 cooling it and arctic cooling mx-2 for my paste.

I was running a stress test with OCCT and had 2 other temperature programs open as well (Realtemp and coretemp). While OCCT and Coretemp came pretty close to agreeing (+/- 2 degrees per core or so) Realtemps is several degrees cooler than the other two. I hit the test senor button and it showed it was within the boundaries.

So who do I believe?

Idle:
http://img383.imageshack.us/img383/6295/23955042wj3.jpg

OCCT stress, at it's hottest, per all three programs:
http://img413.imageshack.us/img413/6228/15737881kq6.jpg

 

[Guest]

real temps is on average 5 - 10 degrees cooler than core temp

i'd go with core temp... because even if its not accurate... it will give you a better stable temperature.... like going off of 70c even if its 60 c.... if it says its 70 c but really 60 c... and you think 70 c is too high... and you figure out a way to lower it... it will be better because then the "actual" temperature will be lower than 60 degreees

 

[CompuTronix]

Core Temp uses Tjunction Max 105c for 45 nanometer processors, and 100c for the 65 nanometer G0 processors, which are both wrong. Real Temp uses 95c, which is based upon extensive research, testing, and analysis, and has been proven correct. The author of Core Temp has performed no such testing.

Please read the following thread: Which do i believe? - http://www.tomshardware.com/forum/248039-29-which-believe#t1769260

The author of Real Temp offers his comments near the end of the thread.

Comp

 

[kumquatq3]

Read that other thread, so because my 65nm is a G0 stepping, core temps is no longer accurate? Thanks again.

 

[Shadow703793]

+1 to Comp.

 

[CompuTronix]

Correct. Core Temp is not accurate. Core Temp uses Tjunction Max 100c for the Q6600 G0, which is wrong. Real Temp uses 95c, which is correct. The author of Real Temp has specifically tested the Q6600, and his findings are correct.

 

[kumquatq3]

Thanks again!

 

[CompuTronix]

Glad I could help, but I'm just curious; if you read the other thread, then why were not convinced? Why did you feel that it was necessary to post another thread asking the same question? Did we not fully explain everything, and did we not answer all posible questions?