Thu, Aug 18, 2022 @ 02:00 PM - 03:30 PM
Ming Hsieh Department of Electrical and Computer Engineering
Conferences, Lectures, & Seminars
Speaker: Daniel Brunner, Institut FEMTO-ST, Université Bourgogne Franche-Comté CNRS UMR 6174, Besançon, France.
Talk Title: Towards Scalable Photonic Neural Networks with (3+1)d Integrated Optics
Abstract: Integrated photonic architectures have the potential to revolutionize neural network computing. However, conventional 2D lithography strongly limits the size of integrated photonic neural networks due to fundamental scaling laws. This is of particular importance since scalability to large network sizes proofs to be of crucial importance for neural network computing performance. We want to overcome this problem by integrating neural networks using 3D printed photonic waveguides. For that, we demonstrate complex 3D multimode waveguide networks based on polymer waveguides surrounded by air. Furthermore, we recently developed a (3+1)D direct laser writing technique where we dynamically and locally control the writing power in order to realize single mode step or graded index waveguides.
Biography: Prof. Daniel Brunner is a CNRS researcher with the FEMTO-ST, France. His interests include novel computing using quantum or nonlinear substrates with a focus on photonic neural networks. He was received several University and the IOP's 2010 Roys prize, the IOP Journal Of Physics:Photonics emerging leader 2021 prize and an ERC Consolidator grant in 2022. He edited one Book and three special issues, has presented his results 45+ times upon invitation and has published 50+ scientific articles.
Audiences: Everyone Is Invited
Contact: Marilyn Poplawski