Information is Physical

Abstract

Thermodynamics arose in the 19th century out of the attempt to understand the performance limits of steam engines in a way that would anticipate all further inventions. Claude Shannon, after World War II, analyzed the limits of the communications channel. It is no surprise, then, that shortly after the emergence of modern digital computing, similar questions appeared in that field. It was not hard to associate a logic gate with a degree of freedom, then to associate kT with that, and presume that this energy has to be dissipated at every step. Similarly, it seemed obvious to many that the uncertainty principle, $\mathrm{\Delta E\Delta t\sim \hslash ,}$ could be used to calculate a required minimal energy involvement, and therefore energy loss, for very short $\mathrm{\Delta t.}$