Modeling Silicon-Photonic Neural Networks under Uncertainties

TL;DR

This paper investigates how fabrication and thermal uncertainties affect the accuracy of silicon-photonic neural networks, revealing significant potential accuracy loss due to these uncertainties in realistic scenarios.

Contribution

It provides the first comprehensive hierarchical analysis of uncertainty impacts on MZI-based SPNNs, highlighting their effect on classification accuracy.

Findings

Uncertainties can cause up to 70% accuracy loss in SPNNs.

Impact varies based on device location and phase tuning.

Simulations show even mature fabrication processes are affected.

Abstract

Silicon-photonic neural networks (SPNNs) offer substantial improvements in computing speed and energy efficiency compared to their digital electronic counterparts. However, the energy efficiency and accuracy of SPNNs are highly impacted by uncertainties that arise from fabrication-process and thermal variations. In this paper, we present the first comprehensive and hierarchical study on the impact of random uncertainties on the classification accuracy of a Mach-Zehnder Interferometer (MZI)-based SPNN. We show that such impact can vary based on both the location and characteristics (e.g., tuned phase angles) of a non-ideal silicon-photonic device. Simulation results show that in an SPNN with two hidden layers and 1374 tunable-thermal-phase shifters, random uncertainties even in mature fabrication processes can lead to a catastrophic 70% accuracy loss.