Optimization of 3x3 neuromorphic photonic network for programmable Boolean operations
Andrea Marchisio, Ali Cem, Yunhong Ding, Vittorio Curri, Andrea Carena, Francesco Da Ros, Paolo Bardella · 2024
We propose a methodology to analyze a 3×3 Mach-Zenhder-based neuromorphic optical network used as a programmable logic gate. The investigated approach starts from the electromagnetic simulation of the integrated optical elements, then moves to the description of the thermal heaters including thermal cross-talk, and finally addresses the definition of the logical levels.