Scalable Superconductor Ising Machine for Combinatorial Optimization Problems
Beyza Zeynep Ucpinar, Sasan Razmkhah, Mehdi Kamal, Massoud Pedram · 2024
Complex combinatorial optimization problems serve as the foundation for various real-world applications. The time required to identify the optimal solutions to these problems escalates dramatically as the problem grows. Nevertheless, transforming these problems into another NP-complete problem, like the Ising model with a representation as a physical phenomenon, can be efficiently approximated. In this work, we use a system of bistable Josephson parametric oscillators (JPO) as artificial spins to realize the Ising model on a superconductor fabric. By integrating these JPOs based on the Lechner, Zoller, and Hauke (LHZ) architecture, we design a superconductor-based scalable Ising machine (IM). We develop a framework to automatically create the IM circuit and tune its parameters. The circuit functionality is assessed by simulating an IM designed for solving four-, six-, and ten-node unweighted Max-Cut problems.