Rigorous dynamic model of a silicon ring resonator with phase change material for a neuromorphic node

Area of Science:

• Photonics
• Neuromorphic Computing
• Materials Science

Background:

• Integrated photonics offers low latency and energy-efficient connectivity for neuromorphic computing.
• Phase change materials (PCMs) enable non-volatile optical memory in photonic circuits, crucial for artificial neural networks.
• Plasticity in artificial neural networks, inspired by the brain, requires adaptable non-volatile weighting elements.

Purpose of the Study:

• To develop a computationally efficient dynamical model of a silicon ring resonator (RR) enhanced by a phase change material (Ge2Sb2Te5 - GST).
• To accurately model the optical behavior of GST within a silicon RR, considering absorption and asymmetry.
• To enable modular integration into system-level photonics simulations.

Main Methods:

• Extended existing dynamical models of silicon RRs and GST thin films.
• Modified optical equations to incorporate GST's high absorption and asymmetry in the RR.
• Included silicon nonlinear effects (free carriers, temperature) and GST phase change dynamics.
• Restructured equations for modular use in commercial photonics simulation software.

Main Results:

• The model demonstrates enhanced speed and energy efficiency for memory operations by factors of 6-10.
• Optical contrast from GST phase change increased over 10x by leveraging RR resonance, with increased optical loss.
• Nonlinear dynamics in silicon RR networks achieved strong contrast while maintaining energy efficiency.

Conclusions:

• The developed model accurately captures the behavior of GST in silicon RRs for neuromorphic applications.
• Design parameter exploration shows significant improvements in memory operation speed and energy efficiency.
• The study highlights the potential of resonant structures for enhancing optical contrast in PCMs for photonic neural networks.