4 ms·

Wasn't it a consequence of Maxwell's Demon like that as soon a you revert any operation theoretically there will be no energy consumed? Because cleaning up mem

• DennisP 15y ago
  There are people working on reversible computing hardware for that very reason. You still have to consume energy to output information, but that's relatively insignificant. Kurzweil talks about it in The Singularity Is Near, when he examines just how far Moore's Law can go. If we can figure out reversible computing, the upper bound is quite a bit further out, since heat dissipation issues go away. Ask a reversible computer a yes-no question that takes enormous computation, and it will consume no more energy than a rock, until at last you have to pay for your one bit of output.
  • simoncion 15y ago
    Have you any information published by the folks who are working on this task?
    • DennisP 15y ago
      No, sorry. Aside from Kurzweil, I read an article a year or two ago about someone who'd made a bit of a breakthrough, figuring out how to break a large computation into a lot of smaller reversible chunks. But I don't have the link. Edit: I seemed to remember that IBM had done some work on it, and found this paper: http://www.research.ibm.com/people/b/bennetc/bennettc19734c533842.pdf http://www.research.ibm.com/people/b/bennetc/bennettc19734c5... Then I remembered the name Fredkin, and found this: http://www.aicit.org/jdcta/ppl/18.%20JDCTA4-440054.pdf http://www.aicit.org/jdcta/ppl/18.%20JDCTA4-440054.pdf Fredkin invented the Fredkin Gate: http://en.wikipedia.org/wiki/Fredkin_gate http://en.wikipedia.org/wiki/Fredkin_gate
  • wnoise 15y ago
    And quantum computation must be reversible too, except for the final measurement process.