CONFERENCE PROCEEDINGS
Advanced Search
Home > Proceedings > Volume 12757 > Article
Paper
17 August 2023 A wavelength-multiplexed photonic tensor processor based on Mach-Zehnder modulator
Hongfei Li, Zeyu Lin, Sheng Zhang
Author Affiliations +
Proceedings Volume 12757, 3rd International Conference on Laser, Optics, and Optoelectronic Technology (LOPET 2023); 127571T (2023) https://doi.org/10.1117/12.2690776
Event: 3rd International Conference on Laser, Optics and Optoelectronic Technology (LOPET 2023), 2023, Kunming, China
Abstract
Photonic processors have shown great potential to replace electronic processors due to the high parallelism, low latency, and low power consumption of light. The current direction of integrated optical computing research is focused on realizing vector matrix multiplication in light, which is also called photonic tensor processor. There are two main implementation schemes of photonic tensor processors: coherent and wavelength division multiplexing, and the latter is represented by the micro-ring resonator (MRR) weight bank. However, the MRR is highly sensitive to the manufacturing error, and it takes a long time to calibrate before used for calculation, which makes the realization of a large-scale MRR weight bank a challenge. We propose a novel architecture of wavelength-multiplexed photonic tensor processor based on Mach-Zehnder modulators (MZMs), which can reduce the time and power cost of device calibration We implement a 4×4 optical convolution operator based on the above architecture, which is used to solve the MNIST handwritten digit recognition problem. Prediction accuracy of 91.72% is achieved compared with the accuracy 97.58% achieved by a computer.
© (2023) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Citation Download Citation
Hongfei Li, Zeyu Lin, and Sheng Zhang "A wavelength-multiplexed photonic tensor processor based on Mach-Zehnder modulator", Proc. SPIE 12757, 3rd International Conference on Laser, Optics, and Optoelectronic Technology (LOPET 2023), 127571T (17 August 2023); https://doi.org/10.1117/12.2690776
ACCESS THE FULL ARTICLE
ORGANIZATIONAL
Sign in with credentials provided by your organization.
INSTITUTIONAL
Select your institution to access the SPIE Digital Library.
SELECT YOUR INSTITUTION
PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.
PERSONAL SIGN IN
No SPIE Account? Create one
PURCHASE THIS CONTENT
SUBSCRIBE TO DIGITAL LIBRARY
50 downloads per 1-year subscription
Members: $195
Non-members: $335 ADD TO CART
25 downloads per 1 - year subscription
Members: $145
Non-members: $250 ADD TO CART
PURCHASE SINGLE ARTICLE
Includes PDF, HTML & Video, when available
Members: $17.00
Non-members: $21.00 ADD TO CART
PROCEEDINGS
 5 PAGES
DOWNLOAD PAPER SAVE TO MY LIBRARY
GET CITATION
Advertisement
Advertisement
RIGHTS & PERMISSIONS
Get copyright permission  Get copyright permission on Copyright Marketplace
KEYWORDS
Neural networks
Matrix multiplication
Optical computing
Power consumption
Wavelength division multiplexing
RELATED CONTENT
Subscribe to Digital Library
Receive Erratum Email Alert
Back to Top