5 temperature program test

[drjulian]

Discussion about varying nos.from temps programs led to this test.
I ran

(1) CPUID Hardware Monitor (HWM)
(2) Core Temp 96.1 .........(CT)
(3) Sensors View Pro........(SVP)
(4) Speed Fan 4.33 ........(SF)
(5) INTEL Desk Utilities.....(IDU)
............................................................all at once.

I used PC Wizard 2008 Windows Benchmark as stressor. .SVP shows MUCH higher stress than SuperPi

My system: Gateway c2d E6400(2.13ghz) 2gb generic RAM(533hz?) 400gb Hitachi HD XFX 8600 GTxxx Arctic Freezer Pro CPU Cooler VISTA Ultimate ....

HWM,CT,SVP kept almost identical nos., about 15c (+/- 2)......HIGHER ! .... than SF,IDU (which were also almost identical).....

It is my belief that Intel engineers who fabbed this chip,should be the baseline

Unfortunately,IDU only shows one core temp,but it appears to be (0) the hotter one,and they show numerous MBO temps.
SF has good charts,but I still need to learn more to setup

To me,the most surprising result was to find by far my hottest component was ICH on MBO,reached 80c+ My CPU idled 48-50,maxed 63-65,but I have lots more fan speed (can hardly hear it) if i ever learn how to increase!
................................................................................................... Personally, I will run SF graphs with IDU gauges. ....... Don Julian

Later ...(ArcticPro7 has fins bent to deflect airflow to MBO (stock) , ICH 80c at IDLE! DROPS to 76 at max.)

 

[CMH]

Sorry, but that is way too hard to read.

 

[drjulian]

Thanks for reply.I agree too hard,but the board jams everything together! If I could do WYSWYG (what you see is what you get) it would be better,less abbreviations would help, Tables would help ,too,but I see from your "Post Temps" there's none...Please keep on telling me when you feel I'm wrong,makes me do better....Reading other boards here makes me feel you should be nominated as saint,the many people you've helped....Thanks again,....Don

 

[CMH]

I just use the quick reply... I get an almost what-you-see-is-what-you get there (html tags doesn't show up, but works).

Using enter to create a new paragraph on a new line of thought helps with the reading.

And you're right, there doesn't seem like a tabling system.

And thanks for that compliment . I can get a little bored in the real world.... Besides, I have an interest in cooling/overclocking. I don't really help (or even read) the other forums much. Sometimes I venture into the CPU or Graphics forum, but since those questions don't quite interest me, I don't chase up on them

 

[drjulian]

OK , I edited it,better?

 

[kimsland]

It's very difficult to confirm these senses, as some copy info directly from CMOS and some take info directly from the Motherboard.
In general the best program is the manufactures program hopefully available for download, on each individual hardware component (ie use intel for intel)

There was another thread on this issue (about cpu temp differences on different programs) but you would like a few different senses for a few different hardware components.

After lots of research (googling) it was found that the only one true sensor, was a meter directly attached to the bottom of the CPU (now that's a concern!)

So through a little guestamation (from motherboard reporting values that even the power supply voltage can change) in the end, it is better to use the most commonly used (or manufacture supplied) testing tools. And then just accept it (with tolerance % of who knows!)

Unresolved !

 

[SNGX1275]

You can do a (code) paste (/code) but brackets instead of parenths and then it will preserve the spacing.

 

[CMH]

That is way way way easier to read

I think no matter how much you compare these temps, it still would be pointless until you actually have a calibrated probe in the list as well.

And even then, it will only be true for you, the guy who did this test. All mobos will be different. Same for overclocking, you can't do an overclock, and expect all other motherboards/CPUs to have the same overclock potential.

kimsland: Yes, it is true that the only way to measure the true CPU temp is to have a probe UNDERNEATH the CPU cover, and on the CPU itself. This would be pretty pointless in itself, since placing one there will cause whatever cooling you put on it to be less efficient.

The most practical way would be to put it beside the CPU itself, on the underside of the heatsink. The temps read will be lower though....