Temperature differences on X2's

[rhiridflaidd]

I've just found out that Everest now reports the difference between the temps of each core on a system.

 

http://www.lavalys.com/products/down...?ps=UE&lang=en

 

One of my cores runs 10 degrees hotter than the other, even after replacing my heatsink, and putting down some Artic silver. My core temps have dropped 1-2 degrees after this, and it still hasn't cured, but there's still a 10 degree difference.

 

This strongly suggests means that thie integrated heatsink is poorly assembled, and that if I take a scalpel to the processor, the odds are that my overclock would improve loads, but decapitation runs a risk of having no more CPU. And 939 stuff aren't easy to replace.

 

I now wish I hadn't found this out! The good (cooler) core does 2900mhz, stable.

[ebdoradz]

its normal the core temps arent accurate

 

i have a core @ 15°C and one @ 21 on mine and my overall cpu temp is 30°C

[rhiridflaidd]

But which core falls over first on prime?

 

http://forums.amd.com/lofiversion/index.php/t91541.html

 

seems very interesting.

[kitfit1]

As has been said, differences between the cores is normal with the HIS on. But you can get them closer together by going "lidless". It will be worthwhile and fairly simple to do. It makes it even simpler if you have a HS that bolts on as well. Even under phase it made a big difference in my case. Taking the HIS of i managed to get my cores within 1c of each other.

[rhiridflaidd]

Yep, that's what I'm thinking. My watercooler is a bolt on.

 

The hot core caps out at 2650mhz. The cool core can prime at 2900 mhz. I guess i'm asking too much to expect to go up to 2900 but it's tempting, just to see....

 

But I'm afraid of dammaging the 939 preocessor. Because finding a replacement will be really difficult, thanks to AMD dropping 939.