A new computing architecture using Ising spin model implemented on FPGA for solving combinatorial optimization problems

Abstract

Recently, the new computing architecture using Ising spin model has been attracting considerable attention. It is well known that the Ising spin model represents the physical properties of ferromagnetic materials in terms of statistical mechanics. In this model, the spin states are varied in order to minimize the system energy automatically, by the interaction between connected adjacent spins. The new computing scheme maps combinatorial optimization problems based on Ising model and solves these problems by using ground state search operations exploiting its convergence property. In this report, a new computing architecture using Ising spin model was implemented using field-programmable gate array (FPGA), and Ising computing using FPGA was investigated to solve combinatorial optimization problems.