Publication date: 17th July 2025
Photonic Synaptic Devices for Ultra-Low-Power Neuromorphic Computing
Vellaisamy A. L. Roy
School of Science and Technology, Hong Kong Metropolitan University, Hong Kong
vroy@hkmu.edu.hk
The advancement of neuromorphic artificial intelligence (AI) systems marks a paradigm shift toward computing architectures that emulate the brain’s remarkable concurrency and ultra-low power efficiency. Central to this vision are photonic synaptic devices, which leverage the unique advantages of light to replicate synaptic functions with unprecedented speed and minimal energy dissipation. Unlike traditional electronic counterparts, these devices enable massively parallel information processing, high bandwidth, and scalable integration, addressing the growing demands of next-generation AI applications. By uniting advanced material platforms with innovative photonic design, neuromorphic hardware can transcend the conventional energy–performance wall, paving the way for sustainable, ultra-high-performance computing. This approach not only facilitates real-time learning and adaptive behaviour in artificial systems but also opens new pathways for energy-efficient, large-scale neural network implementation.
Keywords: Neuromorphic Computing, Photonic Synaptic Devices, Ultra-Low Power, Parallel processing, Artificial Intelligence
Hong Kong Metropolitan University