RESEARCH SPOTLIGHT
Hi, I am
Mingyang Gao
HOST INSTITUTE
Femto-ST, CNRS
PROJECT TITLE
Deep all-optical neural network based on complex LA-VCSELs at GHz bandwidths
KEY WORDS
All-optical neural networks
Photonic computing
Large-area VCSELs (LA-VCSELs)
Neuromorphic photonics
High-speed optoelectronics
RESEARCH AIM
My research aims to build scalable, integrated all-optical neural network platforms so that photonic computing can be more easily tested, developed, and applied in real-world scenarios.
RESEARCH OVERVIEW
My project explores how light can be used to perform neural-network calculations at very high speed. Building such a system requires much more than optical components alone: it brings together lasers, detectors, high-speed electronics, mechanical design, control hardware, data processing, and software. I am developing and integrating these elements into a complete experimental platform for testing optical neural networks in a flexible and reliable way. The long-term goal is to make these systems easier to evaluate, scale, and apply, helping bridge the gap between laboratory demonstrations and potential real-world photonic computing applications.
RESEARCH CHALLENGE
The main challenge is to turn high-speed all-optical neural networks from laboratory-scale demonstrations into scalable and reliable computing systems. This requires not only suitable photonic architectures, but also the integration of high-speed detection, data acquisition, control electronics, signal processing, mechanical design, and software. My research therefore focuses on overcoming these system-level bottlenecks so that optical neural networks can be characterized, scaled, and eventually explored for practical applications.
RESEARCH INNOVATION
RESEARCH IMPACT
What excites me most is the possibility that my research could help move photonic neural networks beyond laboratory demonstrations and closer to practical industrial applications. I find it particularly motivating to work on the system-level challenges that may eventually make these technologies more usable, scalable, and relevant outside the research laboratory.
RESEARCH SUMMARY
My research focuses on developing deep all-optical neural networks based on complex large-area vertical-cavity surface-emitting lasers (LA-VCSELs) operating at GHz bandwidths. The project is highly interdisciplinary, combining photonic neural-network architectures with high-speed electronics, mechanical design, embedded control, signal processing, and software integration. Current work includes the development of a custom high-speed photodetector array and a multichannel acquisition system, together with the supporting control and integration infrastructure required for optical neural-network experiments. A key objective is to build a flexible, reliable, and scalable experimental platform for systematic characterization of performance, stability, and processing speed. By integrating optical, electronic, mechanical, and software subsystems into a complete solution, the project aims to lower the practical barrier to using all-optical neural networks and to support the exploration of future real-world photonic computing applications.
RESEARCH VISUAL
RESEARCH JOURNEY
RESEARCH SECONDMENT
Secondment 1
Host organisation: Akhetonics GmbH
Country: Germany
Duration: 2 months
Main objectives:
To investigate high-speed digital system architectures for handling the large data throughput generated by photonic computing systems, with particular emphasis on high-speed data buffering, transfer, and real-time processing. The secondment will also provide experience in integrating electronic data-handling hardware with high-speed photonic systems in an industrial environment.
Skills expected to gain:
High-speed digital system data buffering and transfer architectures, real-time data processing, and practical experience in industrial photonic system development.
Secondment 2
Host organisation: Aston University – Aston Institute of Photonic Technologies (AiPT)
Country: United Kingdom
Duration: 3 months
Main objectives: (Provisional)
To explore neuromorphic photonic processing for high-speed optical communications, with particular interest in evaluating optical neural networks for communication-related signal-processing tasks. The work is expected to include performance assessment, numerical or experimental studies, and investigation of how photonic neural-network architectures can be applied to high-throughput optical communication systems.
Skills expected to gain:
Optical communication system modelling and characterization, photonic signal processing, neuromorphic computing for communication applications, and experience in applying optical neural networks to practical high-speed information-processing tasks.
LEARNING GOALS
RESEARCH BENEFITS
CAREER DEVELOPMENT
MORE THAN PHD
Why did you decide to join POSTDIGITAL+ ?
What do you think is the biggest advantage of being part of the International Doctoral Network?
The biggest advantage is the opportunity to experience different research environments through international collaboration and secondments. This broadens both technical and cultural perspectives, while making it easier to exchange expertise, build long-term collaborations, and connect academic research with industrial applications.
What skill have you developed the most so far?
What are your career ambitions after completing your doctorate?
After completing my doctorate, I aim to pursue a career in industry, ideally in a research and development role related to photonics, high-speed electronics, optoelectronic systems, or advanced computing hardware. I would like to work on technologies that combine strong scientific content with practical engineering and real-world applications.
What inspired you to pursue this research Area?
What surprised you most since starting your PhD?