Sub-milliwatt threshold power and tunable-bias all-optical nonlinear activation function using vanadium dioxide for wavelength-division multiplexing photonic neural networks

TL;DR

This paper introduces a low-power, tunable all-optical nonlinear activation function using vanadium dioxide for wavelength-division multiplexing photonic neural networks, demonstrating potential for high-speed, energy-efficient neural computation.

Contribution

It presents a novel VO2-based waveguide device with sub-milliwatt threshold power and tunable bias, optimized for WDM photonic neural networks, with detailed simulations and performance evaluation.

Findings

Sub-milliwatt activation threshold achieved

Device footprint of 5 micrometers

Temporal rise time as low as 5 microseconds

Abstract

The increasing demand for efficient hardware in neural computation highlights the limitations of electronic-based systems in terms of speed, energy efficiency, and scalability. Wavelength-division multiplexing (WDM) photonic neural networks offer a high-bandwidth, low-latency alternative but require effective photonic activation functions. Here, we propose a power-efficient and tunable-bias all-optical nonlinear activation function using vanadium dioxide (VO2) for WDM photonic neural networks. We engineered a SiN/BTO waveguide with a VO2 patch to exploit the phase-change material's reversible insulator-to-metal transition (IMT) for nonlinear activation. We conducted numerical simulations to optimize the waveguide geometry and VO2 parameters, minimizing propagation and coupling losses while achieving a strong nonlinear response and low-threshold activation power. Our proposed device features a sub-milliwatt threshold power, a footprint of 5 $\mu$m, and an ELU-like activation function. Temporal dynamics show a rise time as low as 5 $\mu$s. Moreover, the bias of our device could be thermally tuned, improving the speed and power efficiency. On the other hand, performance evaluations using the CIFAR-10 dataset confirmed the device's potential for convolutional neural networks (CNN). Our results show that a hybrid VO2/SiN/BTO platform could play a prominent role in the path towards the development of high-performance photonic neural networks.