Why can't you make a superconducting microprocessor?

its possible, as a physics major i stand behind it. There could be a lot of potential problems with it, and cost is another factor. Another question is why do we still use copper to cool cpu's there are other "better" alternatives.

It could be done however to run a computer 24/7 one needs a cryogenic cooling machine like this
http://en.wikipedia.org/wiki/Hampson-Linde_cycle
http://en.wikipedia.org/wiki/Cryogenics#Production

However it takes an enormous amount of power to run one of these machines.
A small version to cool a CPU would still use a lot of power.

cpu produce a lot of heat. cant "super-conduct" heat-generating object.

The industry has by orders-of-magnitude the most experience with semiconductor transistors that need a certain minimum temperature to function. Basically if a semiconductor material is too cold, not enough electrons get excited into the conduction band (sorta like my wife! ). So it behaves more like an insulator and fails to function at such low temperatures. This is why satellite & space probe electronics need a heater to keep them warm enough to function, if their self-generated heat is insufficient.

The LN2 or even LHe oc's don't cool the CPU down to the point where the transistors stop functioning - they basically just pump away the enormous heat the chip generates. Any CPU that starts out at superconductor temperatures is not likely to function.

There are superconduction switches (Josephson junction devices for one) but the industry has little experience trying to fab a billion of these into a modern CPU - it would take a lot of $$ and time to do this, and currently not worth the effort.

This is similar to why we don't use ternary (three active level) logic instead of binary ON/OFF logic, even though somebody once proved mathematically the most efficient logic possible would use base "e" (2.718) logic states, and 3 is closer to this than 2.

"high temp" superconductors right now only works @ temperatures below 140k. above that temp and they cease to be superconductors

Might you have this proof? I don't doubt it or anything, it just sounds like an interesting read (I'm an engineering major).

There is already one at least. the was an article here in toms about it. It is a question of materials really. Find me a material that is super confctor at room temperature or worse , that is super conductor with a temperature delta from -10º to 45º Celcius. And that doesn't supper from magnetical interference. And that doesn't generate (like every other encapsulated chip) a strong magnetic field. Find a material that is stable and that doesn't react or degrade to a crap load of conditions, and they will do it pretty fast.

They tried bananas, but didn't worked.

Unfortunately, no I don't have it. However I believe it was by Donald Knuth, published in the early '70s. I remember running across it while researching a patent application on ternary logic about 15 years ago. It was more mathematical proof than anything else - interesting at the time but just a diversion from my project, so I scanned it on my lunch break.

Depends on how much the line losses are driving thermals vs how much is from leakage.

That would be sweet. Now, to create high-Gauss nano-magnets on silicon!