All-Optical Neuromorphic Computing: Overcoming Material and Functional Challenges

A number of breakthroughs are still required to unlock the potential of neuromorphic photonics. We foresee that the inclusion of new materials, the development of advanced hybrid integrated schemes, the implementation of new functionalities will enable more efficient sensing, storage and processing on chip, to enable large-scale computation and ultra-low energy consumption, shaping the future of neuromorphic computing.This research topic addresses the open questions and challenges in the development of all-optical neuromorphic computing devices, in relation to both spiking and deep neural networks, with a strong focus on new materials and functionalities that could make computing faster, scalable and energy efficient.The present research topic will focus on the integration of novel materials for sensing, storage and processing functions within single or multiple devices. This topic will examine the limitations in current material choices, and propose potential solutions in terms of new materials and/or new functionalities. By investigating these challenges, the research aims to pave the way for more efficient and versatile all-optical neuromorphic systems. Exploration of the impact of these new functional material systems in complex applications and their exploitation for in-situ training are also part of the topic interests. Topic editors: Qiming Zhang, Ripalta Stabile Source: https://www.frontiersin.org/research-topics/72747/all-optical-neuromorphic-computing-overcoming-material-and-functional-challenges