Optimizing Energy-based Neural Network Training with Coherent Ising Machine
2026-06-08 • Machine Learning
Machine LearningArtificial Intelligence
AI summaryⓘ
The authors used a special kind of physical device called a Coherent Ising Machine (CIM) to train a type of neural network in a way similar to software methods. They improved the training process by adding an optimizer called Adam, which helped the network learn faster and more accurately. They also showed this approach can work well for bigger and more complex networks. Their work suggests that these physical machines could be a new and efficient way to build AI hardware in the future.
Ising machineCoherent Ising MachineEquilibrium PropagationHopfield networkAdam optimizerenergy-based neural networkcombinatorial optimizationconvolutional networkanalog circuitsoptoelectronics
Authors
Chen-Rui Fan, Bo Lu, Zhi-Hong Zhang, Run-Qing Zhang, Jing-Wei Wen, Chuan Wang
Abstract
While Ising machines serve as advanced physical solvers for the Ising model,enabling applications in combinatorial optimization and neural network training,their scalability for large-scale neural networks remains constrained by hardware connectivity limitations and suboptimal training methodologies. In this work,we leverage a Coherent Ising Machine (CIM) to train an energy-based neural network using Equilibrium Propagation, achieving performance comparable to existing software-based implementations. We further enhance the algorithm by integrating the Adam optimizer to solve for the ground state of a Hopfield energy network, significantly improving convergence speed and solution accuracy. Additionally, we demonstrate the scalability of our approach across deeper network architectures and convolutional operations. Our results highlight the potential of CIM dynamics as a scalable platform for training complex neural networks, offering a pathway toward energy-efficient implementations via analog circuits, optoelectronics, or integrated photonics. This work establishes a novel physical framework for next-generation AI hardware development.