Critical memory capacity of Hopfield model implemented in coherent Ising machine

Abstract

The coherent Ising machine (CIM) has attracted attention as one of the most effective Ising computing architectures for solving large-scale optimization problems because of its scalability and high-speed computational ability. The CIM is a non-equilibrium open-dissipative system, so the theories and techniques of classical equilibrium thermodynamics cannot be directly applied to it. Our research group has adapted these theories and techniques to work with the CIM. Here, we focus on an infinite loading Hopfield model, which is a canonical frustrated model of Ising computation. We derive a macroscopic equation to elucidate the relation between critical memory capacity and normalized pump rate in the CIM-implemented Hopfield model.