On 10 Apr 2001, at 20:50, Shane & Amy Sanford wrote:
>
> >So my open question is: Do sudden temperature changes cause physical
> >harm to CPUs, especially those that have only recently been turned
> >on?
>
> Yes it does but.... this depends on your processor & cooling methods
> to a certain extent. It's generally accepted that sudden temp changes
> will slowly degrade your processor over time
Eh? A processor with cracking will either work or not. Slow
degradation is more likely aging by thermal diffusion.
>
> As for as CPU temps on boot up my experience has been that AMD
> thunderbirds & durons heat up very very fast so by the time you boot
> into Windows the core will be running pretty close to the temps it
> will be leveling out at (assuming you are using some form of air
> cooling).
If you have a CPU which turns off an inactive floating-point unit,
you may well be only about halfway between ambient and "normal"
operating temperature by the time the system's finished booting. Even
if the boot time is long compared with the thermal inertia time
constant.
> The temps reported by the BIOS on the other hand takes a
> while to level out because of the way the temps are taken. The reason
> this happens is fairly complicated and beyond the scope of the
> question.
Basically (a) there is thermal inertia - _lots_ in the heatsink, (b)
socketed CPUs usually have the thermistor mounted in the well in the
centre of the CPU socket. So it's measuring the temperature of the
socket well, not the heatsink, and certainly not the chip substrate.
And it's only the substrate temperature that actually matters.
Probably a variable fan controlled by the temperature sensor would
help stabilise the substrate temperature, but it's bound to be
stabilised somewhere close to the max temperature, so what you gain
in reduced cracking damage might be more than offset by accelerated
diffusion ageing.
> So in my
> opinion I wouldn't worry about being "careful" with the CPU right
> after boot up because the actual boot up & shut down are where the
> most damage is done anyway.
I agree, but for a different reason - the way CPUs are increasing in
power year on year, that 1.2GHz processor will be pretty puny in two
or three years, so why worry unduly about preserving its life beyond
that?
Regards
Brian Beesley
_________________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm
Mersenne Prime FAQ -- http://www.tasam.com/~lrwiman/FAQ-mers
