[Cryptography] information, Shannon, and quantum mechanics

[Stephan Neuhaus]

On 2015-02-24 19:26, Watson Ladd wrote:
> No, Denker is right. You can perform a reversible computation with
> zero entropy gain. Even if you erase a bit, that might not cost
> energy, but can cost entropy. See for example
> http://rspa.royalsocietypublishing.org/content/royprsa/early/2011/01/07/rspa.2010.0577.full.pdf.

There's also the Feynman Lectures on Computation, where the author 
tackles questions like these (if I remember correctly, from a 
not-too-thorough reading many many years ago).

If I remember correctly (which I might not!), you could do reversible 
computations with zero energy, in the limit.

Fun,

Stephan