Correspondence Between Ising Machines and Neural Networks

TL;DR

This paper establishes a systematic correspondence between Ising machines and neural networks, enabling neural network computations on Ising hardware at high temperatures with proven success.

Contribution

It generalizes Ising-based computation from ground states to spin averages and introduces a method to implement trained neural networks on Ising devices.

Findings

Correspondence between Ising models and neural networks established

Method to run trained neural networks on Ising hardware demonstrated

Mathematical proof of successful implementation provided

Abstract

Computation with the Ising model is central to future computing technologies like quantum annealing, adiabatic quantum computing, and thermodynamic classical computing. Traditionally, computed values have been equated with ground states. This paper generalizes computation with ground states to computation with spin averages, allowing computations to take place at high temperatures. It then introduces a systematic correspondence between Ising devices and neural networks and a simple method to run trained feed-forward neural networks on Ising-type hardware. Finally, a mathematical proof is offered that these implementations are always successful.