Ok, speaking as the primary alarmist of the bunch... let me tell you it's not that big a deal at all. Nobody is talking about heatsinks falling off CPUs or things going up in smoke...
What started all this is that I ran into a problem with a number of AMD machines, mostly in three groups that were soaking up an inordinate amount of my time doing heat related service calls. Because this was strictly an AMD thing, I made a financial decision not to sell AMD based systems. I still own two AMD systems and like them just fine. It has however become rather exaggerated as people leapt to the fore defending their love of AMD and taking swipes at my competence as a technician. I mean really... how dare I discuss a technical issue on a hardware oriented BBS
Methinks the problem is that most people here aren't used to seeing discussion of repair and engineering issues and don't appreciate that one can discuss a problem with a device without disliking the device itself.
The problem, as it turned out, was that these machines were being jarred almost daily by cleaning crews who were picking them up to dust under the fronts of them and then just dropping them the last half inch or so to avoid pinching their fingers. It turns out the offices where the problems were coming up were all using the same cleaning company. They've now been replaced and the problem seems to have moved on with them.
Still, when a half inch drop can disturb a heatsink enough that I have to go out to a customer's office and re-install it, there is a real problem there. That's way too delicate for commercial use.
As I looked further into the problem I did discover some minor inadequacies in the way AMDs handle heat ... nothing disastrous, mind you, but (I thought) enough to warrant comment. Basically it's a mechanical problem of keeping the heatsink stable on top of the cpu's heat island when the top of the heat island itself is not always flat... all the XP processors I have here are slightly domed, only about .05mm, but that's enough to allow some movement, especially when jarred.
Needless to say keeping something from moving on a domed surface is not done with a single point of pressure from a spring clip. So the problem is purely mechanical... the electronics of the AMD CPU are just fine and were it not for this problem I would still be selling AMD in preference to Intel.
The symptoms of movement are fairly easy to recognize; the CPU temperature will creep up slowly over time. Or, if jarred, it will jump about 5c or 10c over night. In a few cases, with machines that are in heavy use, this could (and did) become a problem since it may triggering the system's heat alarms causing abrupt shutdowns. People were losing half and whole day's worth of data processing when the systems simply switched themselves off with no warning. This won't be much of a problem for a homeowner but in an office where the company's well being is on the line it cannot be tolerated.
But, as it turns out there is a simple solution.. I've had some small stabilizer clips custom made and am presently running around putting them in all my AMD systems. This seems to clear the problem right up. For you the solution, should you run into this, would be to purchase one of the readily available non-conductive shims. (<A HREF="http://www.casecooler.com/unnonmicshim.html" target="_new">http://www.casecooler.com/unnonmicshim.html</A>)
My mistake, in terms of the politics of this BBS was thinking I could discuss this with other knowledgeable people without getting my ass burned for daring to find fault with their beloved AMD processors.
--->It ain't better if it don't work<---