"Burn-in" temperature drop????

[chaosdriven]

Hi all!

Last night I finally nudged up my default core voltage by 0.1(1.80->1.89), and set the system to 1550(10*155). I upped the voltage because the system would freeze up at various combinations at about 1550. Well, the voltage increase allowed the system to operate- but at the cost of increased CPU temps. At 1528 my load temps were about 52-53c. I watched with a sense of dread as my temps went up to 58-59c under load, despite this, I let the system run overnight. Now, to my surprise, I see the temps trending down.... currently back down to the 54-55c range with no changes to the system or ambient temperature that I can appreciate.

Is this normal for the "burn-in" process? :confused:

[Hugovsky]

Thats what happened to me too. I think that doing burn in makes thermal paste/chip/cooler work even closer and better...

[Pitspawn]

Yeah, burning in does affect temperatures with bad thermal compound. I dont think it works with Arctic Silver I and II though.

Burning in is actually quite dangerous though because you put your rig under way more strain then you are used to.

[Colin]

The particles in the thermal grease are aligned by the vibration of the HSF or water flowing through a water block. All thermal compounds settle in. I typically see a 2C temp drop with Arctic Silver after a couple of days.

[chaosdriven]

I am using Artic Silver though....

[mhamm]

I had same thing happen to me. Im not complaining :) I'm all for it.,

[Myself and I]

Just testing to see if the board works ok

[silent bob]

I do a burn in til my
temps under full load =my starting idle temps
that is usually 2 days for me then I do bump up the fsb a liitle more after that and it has been a plan that works for me DONT TRY IT IF YOU ARE NOT SURE OR DONT FEEL SAFE unless you want in the KEYRING Club so do at your own risk