Next Article in Journal
A Study of Predictive Models for Early Outcomes of Post-Prostatectomy Incontinence: Machine Learning Approach vs. Logistic Regression Analysis Approach
Next Article in Special Issue
An Integrated Photorefractive Analog Matrix-Vector Multiplier for Machine Learning
Previous Article in Journal
Analytic Binary Alloy Volume–Concentration Relations and the Deviation from Zen’s Law
Previous Article in Special Issue
Neuro-Inspired Computing with Spin-VCSELs
 
 
Article

Photonic Integrated Reconfigurable Linear Processors as Neural Network Accelerators

1
Scuola Superiore Sant’Anna, 56124 Pisa, Italy
2
Department of Information Engineering, University of Pisa, 56122 Pisa, Italy
3
Department of Electrical and Computer Engineering, McGill University, Montreal, QC 3480, Canada
4
CNR-IEIIT, National Research Council of Italy, 56122 Pisa, Italy
*
Author to whom correspondence should be addressed.
Academic Editors: Ilaria Cristiani and Ripalta Stabile
Appl. Sci. 2021, 11(13), 6232; https://doi.org/10.3390/app11136232
Received: 3 May 2021 / Revised: 16 June 2021 / Accepted: 28 June 2021 / Published: 5 July 2021
(This article belongs to the Special Issue Photonics for Optical Computing)
Reconfigurable linear optical processors can be used to perform linear transformations and are instrumental in effectively computing matrix–vector multiplications required in each neural network layer. In this paper, we characterize and compare two thermally tuned photonic integrated processors realized in silicon-on-insulator and silicon nitride platforms suited for extracting feature maps in convolutional neural networks. The reduction in bit resolution when crossing the processor is mainly due to optical losses, in the range 2.3–3.3 for the silicon-on-insulator chip and in the range 1.3–2.4 for the silicon nitride chip. However, the lower extinction ratio of Mach–Zehnder elements in the latter platform limits their expressivity (i.e., the capacity to implement any transformation) to 75%, compared to 97% of the former. Finally, the silicon-on-insulator processor outperforms the silicon nitride one in terms of footprint and energy efficiency. View Full-Text
Keywords: photonic integrated circuit; photonic neural network; optical signal processing photonic integrated circuit; photonic neural network; optical signal processing
Show Figures

Figure 1

MDPI and ACS Style

De Marinis, L.; Cococcioni, M.; Liboiron-Ladouceur, O.; Contestabile, G.; Castoldi, P.; Andriolli, N. Photonic Integrated Reconfigurable Linear Processors as Neural Network Accelerators. Appl. Sci. 2021, 11, 6232. https://doi.org/10.3390/app11136232

AMA Style

De Marinis L, Cococcioni M, Liboiron-Ladouceur O, Contestabile G, Castoldi P, Andriolli N. Photonic Integrated Reconfigurable Linear Processors as Neural Network Accelerators. Applied Sciences. 2021; 11(13):6232. https://doi.org/10.3390/app11136232

Chicago/Turabian Style

De Marinis, Lorenzo, Marco Cococcioni, Odile Liboiron-Ladouceur, Giampiero Contestabile, Piero Castoldi, and Nicola Andriolli. 2021. "Photonic Integrated Reconfigurable Linear Processors as Neural Network Accelerators" Applied Sciences 11, no. 13: 6232. https://doi.org/10.3390/app11136232

Find Other Styles
Note that from the first issue of 2016, MDPI journals use article numbers instead of page numbers. See further details here.

Article Access Map by Country/Region

1
Back to TopTop