Programmable 200 GOPS Hopfield-inspired photonic Ising machine.

Abstract

Ising machines offer a compelling approach to addressing NP-hard problems¹, but physical realizations that are simultaneously scalable, reconfigurable, fast and stable remain elusive. Quantum annealers, such as D-Wave's cryogenic hardware, target combinatorial optimization tasks, but quadratic scaling of qubit requirements with problem size limits their scalability on dense graphs². Here we introduce a programmable, stable, room-temperature optoelectronic oscillator (OEO)-based Ising machine with linear scaling in spin representation. Inspired by Hopfield networks³, our architecture solves fully connected problems with up to 256 spins (65,536 couplings) and >41,000 spins (205,000+ couplings) if sparse. Our system makes use of cascaded thin-film lithium niobate (TFLN) modulators, a semiconductor optical amplifier (SOA) and a digital signal processing (DSP) engine in a recurrent time-encoded loop, demonstrating potential >200 giga operations per second (GOPS) for spin coupling and nonlinearity. This platform achieves the largest spin configuration in an OEO-based photonic Ising machine, enabled by high intrinsic speed. We experimentally demonstrate best-in-class solution quality for max-cut problems of arbitrary graph topologies (2,000 and 20,000 spins) among photonic Ising machines and obtain ground-state solutions for number partitioning⁴ and lattice protein folding⁵-benchmarks previously unaddressed by photonic systems. Our system uses inherent noise from high baud rates to escape local minima and accelerate convergence. Finally, we show that embedding DSP-traditionally used in optical communications-within optical computation enhances convergence and solution quality, opening new frontiers in scalable, ultrafast computing for optimization, neuromorphic processing and analogue artificial intelligence.