Hybridization of Spin Decision Logics for Ising Machine with Logic Circuits

Recent advances in the field of natural computing have boosted the hope that complicated problems can be solved efficiently on the heuristics applying natural phenomena. In particular, it is expected that Ising spin model implemented with CMOS circuits solves combinatorial optimization problems more effectively and conveniently at room temperature, and thus, it is potentially achievable for practical use. So far, we have developed Ising machine that can be implemented with simple logic circuits consisting of XNOR gate utilized as spin decision logics for finding the ground state. In this study, we propose the hybrid-type spin decision logic (HDL), a combination of two conventional methods. HDL can find the global optimal solution of combinatorial optimization problems, that is impossible to solve with conventional spin decision logics.