Introduction

Abstract

In this thesis, we construct a general theory of thermodynamics of information processing. The background of this research is the recently developed field of nonequilibrium statistical mechanics and quantum and classical information theories. These theories are closely related to the modern technologies used to manipulate and observe small systems; for example, macromolecules and colloidal particles in the classical regime, and quantum-optical systems and quantum dots in the quantum regime. First, we generalize the second law of thermodynamics to the situations in which small thermodynamic systems are subject to quantum feedback control. Second, we generalize the second law of thermodynamics to the measurement and information erasure processes used in the demon's memory. Third, we generalize the nonequilibrium equalities such as the fluctuation theorem and the Jarzynski equality to classical stochastic dynamics in the presence of feedback control. In these results, thermodynamic quantities and information contents are treated on an equal footing. Moreover, the obtained inequalities and equalities are model-independent, so that they can be applied to a broad range of information processing applications. Our findings could be called the second law of "information thermodynamics".