Temperature Comparisons may not me valid

One of the most common questions in the forums is what temperature are you running at and if your cpu is running to hot and people comparing temps. I have some info to share with you guys. I took 3 Pentium D 805. Ran them all in the SAME EXACT setup, repeated it twice and got some interesting results. All 3 were overclocked to 3.4GHz, Same cooler and motherboard were used. Now get this all 3 cpu had different load temps. Room temp was the same in all tests. CPU voltage was 1.325v in bios for all cpus
CPU 3: 67C

This test illustrates that we can have the same cpu and the same exact conditions and get different load temps, the difference in temps was quite surprising. 52C vs. 62C vs. 69C. I was expecting maybe 2-3C but 15C.
So when comparing temps with other forum members, take them with a grain of salt.

So really we can't draw any conclusions because in order to do wo we would have to compare your temperatures.

and you know what? there are no 2 identical cpu's, even if they are labeled the same.

yep, thats why you have to test everything out yourself. every mobo is different, every stick of ram, every processor. That's why manufacturers only guarantee stock operation.

The "real" way of getting CPU temp !

Want to know the ultimate truth ?

Just follow Joe's method in his excellent article on Motherboard Test Platform , its the best in the world imho, heck, even Intel's engineer are using this method as their "standard procedure" even on latest C2D processor generation (read it on Intel's own whitepaper) !

A glance what is needed to gain that truth ! Of course, no pain no gain, you'll have to go thru the hassle and right tools though.


Courtesy of www.overclockers.com

2 QX's

I've got two QX's and they both run within 4 degrees of each other with the same setup. I'll agree though, there's a lot of variance between procs.

Your taking the old P4's though. Take a C2D and tempatures are nearly identical. Out of the 4 C2D's I've had, 2 sets where matched nearly identically for temps. 2 of them ran hotter then the other 2 but otherwise they where within 1C of each other if not less since we don't read point anything.

there is statistical deviation to all production procedures. it's just a fact of life and nature. Everything is designed with tolerances...sometimes they are very tight tolerances and sometimes they are loose.