Point of deminishing returns...

[Kryten]

On my way to work this morning I started thing about pelts and their overall efficiency, on a reasonably slow CPU i.e pentium 2 / 300 , which obviously wont produce the same heat as a say 1.4 gig Athlon, is the efficiency of the same size pelt reduced, i.e using the same pelt on both CPU's
the reason I ask is because if there is more heat going in, it must have to get rid of that heat plus any potential difference, if you start with a lower heat it must dissapate easier.
just interested in your thoughts....

[13oots2]

Doesn't heat work on the inverse square law? It's about nearly 20 years since I did any physics, but I seem to remember to lower by 1C would take half the energy of lowering by 2C and 3C would take 4* the energy. That is why you could never achieve absolute zero, it would take an infinite amount energy.
I would think that your efficiency would take a bigger hit than you would expect. Then again my memory could be playing tricks on me. ???

[13oots2]

I'm slowly getting there, the closer a given temperature to the ambient enviroment the harder it is to cool. Tests have actually proven that water of nearly 100 C, when placed in a freezer will freeze quicker than that of water of room temperature. I still can,t work that out myself, but have seen the experiment carried out on a science program.
No wonder I went into Graphic art, not thermodynamics it's too weird

[Kryten]

looks like I have a few people scratching their heads now :)
which leads me to ask this question, if it requires more effort to keep a hotter chip cool, does this mean you need a larger pelt just to do the same job?