Computational Heating
This might be the wrong time of the year to be thinking about this (at least in the northern hemisphere), but this old idea of mine (don’t know if it has been proposed before) has come to mind by the pretty hot (in all senses) MacBook Pro sitting on my lap. If modern computational devices produce so much (waste and wasted) heat, and considering that in some parts of the world electric heating is used extensively, why not combine the two?
Instead of having regular electric heaters to warm up your house and your water, with the decrease in cost of processing capacity, you could have devices that doubled as heaters and processing nodes – processing capacity that you could use for yourself, sell, or donate to projects such as SETI@Home. These devices would be plugged to the mains for power, and have some kind of wireless connectivity. For now, computational heating surely is more expensive (in terms of hardware, not energy) than purely resistive heating, but who knows what the future brings? In a few years time you might be able to go down to the hardware store, and buy a heater (either a space heater or a water heater) with capacity measured both in watts and teraflops!
[...] just read this post on the New Scientist Tech blog, it reminded me of this post I wrote one year and a half ago. Should I be in [...]
I think that’s a great idea! Rather than fight the heat, use the heat. I alawsy wondered why so much heat was being wasted actually. A stirling engine can turn much of that heat inot electrivity after all.