Cover Image

Noise-resilient photonic analog neural networks

The explosion of generative artificial intelligence (AI) has led to an unprecedented demand for AI accelerators. Photonic computing holds promise in this direction, offering speedups in bandwidth and latency. However, photonic integrated circuits (PICs) and their periphery input/output (I/O) compone...

Full description

Saved in:
Bibliographic Details
Main Authors: Varri, Akhil (Author) , Brückerhoff-Plückelmann, Frank (Author) , Dijkstra, Jelle (Author) , Wendland, Daniel (Author) , Bankwitz, Julian Rasmus (Author) , Agnihotri, Apoorv (Author) , Pernice, Wolfram (Author)
Format: Article (Journal)
Language:English
Published: 15 November 2024
In: Journal of lightwave technology
Year: 2024, Volume: 42, Issue: 22, Pages: 7969-7976
ISSN:1558-2213
DOI:10.1109/JLT.2024.3433454
Online Access:Verlag, lizenzpflichtig, Volltext: https://doi.org/10.1109/JLT.2024.3433454
Verlag, lizenzpflichtig, Volltext: https://ieeexplore.ieee.org/document/10609496
Get full text
Author Notes:Akhil Varri, Frank Brückerhoff-Plückelmann, Jelle Dijkstra, Daniel Wendland, Rasmus Bankwitz, student member, IEEE, Apoorv Agnihotri, and Wolfram H.P. Pernice
Description
Summary:The explosion of generative artificial intelligence (AI) has led to an unprecedented demand for AI accelerators. Photonic computing holds promise in this direction, offering speedups in bandwidth and latency. However, photonic integrated circuits (PICs) and their periphery input/output (I/O) components tend to be noisy due to the nature of analog computing. This can lead to accuracy degradation if not accounted for properly. In this paper, we characterize the typical noise levels present in photonic hardware accelerators for deep neural networks (DNNs). We explore several techniques including knowledge distillation, stability training, and standard Gaussian noise injection to improve the robustness of photonic DNNs. We validate our methods by training a Resnet model on the CIFAR-10 dataset and comparing the simulated test accuracy with different noise levels and image distortions. The robust training techniques discussed in this paper combined with the noise analysis of PICs provide a blueprint for robust photonic AI inference accelerators.
Item Description:Online verfügbar: 25. Juli 2024
Gesehen am 15.05.2025
Physical Description:Online Resource
ISSN:1558-2213
DOI:10.1109/JLT.2024.3433454

Similar Items