Vastly different core temperatures on X2.

[cantankerous]

Hey guys, got an X2 4800+ a couple weeks ago and am noticing something very strange. The temperatures of each core are off by as much as 10-12c from eachother!

These temps were noted using the latest core temp as well as the OSD in the newest Everest 3.50. Is this normal or is something wrong with the contact of one of the cores with the IHS? I had one other person on this forum tell me they were seeing similar readings with their Opteron 165. Can anyone else with an X2 or Opteron check to see with core temp what each core is reading seperately? I can have one say 35c and the other will say 43c+.

I however have no stability issues. I ran orthos for 9+ hours without issue, loop 3dmark01 for hours without issue as well as running dual super pi 32M, SnM, OCCT etc. I am hoping this is just a readout glitch rather than the temps really being that high on one of the cores cause if so I can definately see one of the cores going belly up long before the other.

Any info would be appreciated. Thanks.

[Draxx]

Yeah I have 8°C ish difference between cores, no idea why mind....

[cantankerous]

thanks for the response Draxx. I guess I am not the only one afterall.

Anyone else?

[Hemi]

i have the same on am2 4200. idle about 8c load is when they are almost equal in temp

[RAFH]

I have a 4400 X2 and I get 4 to 10 C differences. I used to be concerned but there's been no releases of the magic smoke nor bolts of the magic lightening so I figure its probably not a problem.

[LagunaX]

May sound like a stupid question, but have all of you guys done the amd x2 patch for windows xp and synchronized both cores to run together for all apps with the notepad usepmtimer addendum? Guess it could happen if the cores were running at separate loads and speeds.

[Draxx]

Yeah running the AMD driver and the Optimizer

[cantankerous]

yup same here but nice idea. it does make sense.

I'm not going to worry anymore about it since it seems lots of people are having this. Perhaps it is a core temp issue and not a processor issue itself.

[spdycpu]

I think it might just be oddities of the sensor. I had a 3700+ San Diego I was running with liquid and everything seemed fine as far as temp sensors go. Popped in a 4400+ CCBWE (ugh!) and the core 0 temp was WAY off, reading about 15-20c too low. It did this with the IHS on with a stock 4800+ HSF and Apogee block. It still did it after 2 mounts of the apogee without the IHS as well. At one point my ambient was ~18c, water temp was probably around 20c and core 0 was showing *6c* while core 1 had a more realistic temp. Definitely something weird going on there.

I quickly got rid of the CCBWE and now have an Opteron 165, both cores are more or less the same temps with and without the IHS both during idle and load. I guess some chips have sensors that are faulty or slightly wrong, if you KNOW you've got the mount on properly then I wouldn't worry about it too much. If someone has bad temps on one core they could try it without the IHS I guess, but, if the thing is reporting sub-ambient temps I'd just leave it be, who knows what the thing is really running.

[Sparky]

I had something similar along with really high coretemps overall. I removed my IHS (nerve wracking doing it the first time) and now my temps are much better and the individual core temps only vary by a couple degrees from each other.