Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Not true & hilarious. Exceeding the temperature threshold will change the electrical properties to the point where the manufacturer can't guarantee the performance. For a CPU the timing and leakage will be wrong to the point where the chip might "crash" but it won't fail. Intel shuts the chip down at 125 C mostly because its critical timings will be very wrong, and also because 125 C is indicative of a horrible cooling system failure which might make the chip even hotter. Remember during soldering the chip gets up to 200 C + without any damage.

Tell that to my E6600 and my Northwood 3.0 which failed from a Arctic Silver 5 drying incident.

Processors WILL fail when pushed past thermal limits. Intel's failsafes are not a guarantee of safety.

If you don't think heat has any effects on processors, maybe you should try overclocking. It's well known that a poorly cooled processor will fail faster than a properly cooled one.
 
@vant, Yeah sorry, I apologise for the wording of that remark.

But your statement was clear: you state "temperatures exceeding the threshold will invoke immediate failure" (my bold).

That I think is fundamentally incorrect. you seem to be saying that if we take a Core2Duo to 106 C, it will immediately fail. That's untrue. 1 C difference won't make much difference (if any) to failure probability or lifetime.

On the other hand, you point out that dramatically exceeding manufacturer's temperature specifications can result in instant death, and you give two good examples of that. I entirely agree. Taking a Core2Duo to, say 140 C, will I expect cause instant death.

We are trying to find some common middle ground:- what happens to CPUs which are hot, but are well below, Intel temperature specifications? For the uMBPs the setpoint of the fans seems to be 90 C. The fan speed changes to keep the CPU at that temperature. The rest of logic board seems to be at around 55 C.

My contention was, and is, the following
- your Mac is unlikely to fail in its useful life
- the majority of failures have causes other than logic board death
- the CPU and the logic board do not get hot enough for heat death to be a significant failure mode
- so only a very small number of logic board deaths are heat related
- logic board heat death is very, very unlikely to limit the useful life of your Mac
-> Therefore - just use your Mac and don't worry about temperatures.

Now, the reason I'm going to all this effort is to try to stop users with limited computer/electronics experience from worrying too much - "OMG my CPU's soooo hot, it's gonna kill my Mac". I would much prefer them to worry about things which can and do kill Macs - liquids, dropping, leaving it in the car at -20 C, etc. Also, instilling good habits like backing up regularly.

As another example, on another thread an experienced electronics person is advocating annually taking the back off the MBP and cleaning the heatsink and fans. For an inexperienced user that seems like Mac murder. The probability of them damaging the computer by doing that are IMHO much higher than the heat death probability. But users worried by the "OMG my CPU's soooo hot" posts might do something like this.
 
More succinctly

I think a lot of people out there are buying coolpads instead of backup drives. I think that's wrong. I am not saying "don't buy a coolpad", I am saying "buy a backup drive today". I am also saying "a coolpad almost certainly will not save your ass, but a backup drive almost certainly will"
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.