The new opty is HOT

T

taylormsj

New Member
Just got my new opteron 165 and man is it hot.

Ive got a zalman cnps 9500 and artic ceramic paste on it and it idles at around 44 - 50 C and gets up to about 58 C after 30 mins of a stress test, this is at stock speeds and also at 2.52 GHz with stock voltage, i added another 40 Mv and it went over 60 C, everest doesnt go higher than 60 C so it could have been high 60's or 70 C :eek:.

Im going to keep it at 2.52 GHz on stock voltage until i can figure out what im going to do

im hoping the thermal burn in time drops it signifcantly.

What do you guys suggest, is this normal for an opteron?

Also, core temp, speed fan and pc wizard all read the core temps as being quite low, in the 30 C's at idle, but speed fan, everest and pc wizard read the CPU temp as being about 10 C hotter than the core temps, the BIOS also reads around 46 - 50 C.

Why is there so much difference in core temp to CPU temps?

Thanks
 
You used ceramic instead of AS-5 with silver particles for one. That would offer the improved heat transfer into the Zalman. AS-5 sees an average of a 200hr. breakin time. You may applied too much of the ceramic type paste there as well.
 
I applied the right amount, ive done it loads before, im going to apply some AC5 once my friends build is done if its still high
 
Certain Toledo core chips are just hot, mine for example. It sucks, but there's not much you can do, besides 60*c load isn't hot enough to do any damage. Running it 24/7 like that will obviously show some detrimental effects in a couple years, but why worry about that now?
 
Core temps are what you look at to be certain. That the cpu temp shows a hotter temp than the core temps probably means sensor malfunction, or something. CPU temp should never be that much hotter than the core temps, it's almost impossible, take it as a fluke.
 
My friend has a Opteron, he said he cools his with a cooler with a lot of heatpipes. I overclocked his from 1.8ghz to 2.5ghz. He has to use case fans like 2 or 3.
 
The newer line of AMD models 939s, 940s, AM2 now see a temp thresholh of about 60-70C max. Temp #2 on SpeedFan is not the core but the board temp. Originally it was written for single core cpus to see that, the board, and hard drive temps. With good cooling inplace the #2 temp should be lower then the cpu temp(#1).
 
The point there was simply not to confuse the cpu-board with cpu-core temps. I wasn't sure if you were familiar with that or not. That can easily throw off someone new to using it. On the AM2 board here the core temp is always considerabliy higher then the cpu temps.

The current temp reported here by SpeedFan is about 24C for the cpu. That's with a 125w max output possible for the 6000+ X2 model. With the SIW tool the sensors read 27C for core #0 and 33C for core #1. The System Info for Windows tool is another one you can try to see what you get with that. http://www.gtopala.com/siw_on_pc_world.html
 
An inadequate and poorly regulated supply would more likely see power drop offs when a godo load was placed on it. Besides a 9500 sitting on the cpu the temps will be seen high if the case in general lacks adequate cooling. Another thing is trying to oc as soon as you install a new cpu and not allowing the normal breakin time for thermal paste or compound used. That will see cpu temps soar at times.
 
Hmm yeh, i over clocked it within 10 minutes of installing the cpu, probably not the best idea

Temps-1.jpg


That is what my temps are like, just turned on my PC. I dont think its a case of bad sensors, i want to know why the core temps are so low compared to the CPU temp, the BIOS also reads the CPU temp as bieng about 48 C
 
Last edited:
Now take a reading with the SIW tool and compare that to what you are seeing with SpeedFan and Everest. The move to oc right when first installing the cpu is always strongly advised against. You want to go easy for the first few days until the paste or compound used starts seeing a certain degree of heat transfer take hold. Otherwise the paste bonds to the bottom of the sink and doesn't adgere to the heat shield or exposed chip of the cpu itself.

On the new build here when the first board quit after only 3 days of use all the Artic Silver was found on the copper base of the Zalman 9700 and none on the heat shield of the AMD model here. It takes an average of about two weeks or 200hrs. of use to see a full breakin period allowing for ocing and placing good loads on the cpu then.
 
You must log in or register to reply here.
Back
Top