Achieving More with Less: Training a Spiking Neural Network of 40.000 Neurons with Rank Order Coding, Leveraging Sparsity
We present a scalable photonic spiking neural network (SNN) with 40,000 neurons. Our SNN emulates important aspects of biological neurons by leveraging sparsity [1] for energy efficiency and speed. Photonics has the potential to overcome CMOS-based SNN limitations by utilizing rapid excitable devices as neurons, low-latency optical communication and high parallelism.
Authors
- Daniel Brunner (ORCID: https://orcid.org/0000-0002-4003-3056)
- Anas Skalli (ORCID: https://orcid.org/0000-0002-7039-926X)
- Ria Talukder (ORCID: https://orcid.org/0000-0001-5103-3003)
Institutions
- Centre National de la Recherche Scientifique (FR)
- Franche-Comté Électronique Mécanique Thermique et Optique - Sciences et Technologies (FR)
Publication Details
- Published
- 2025-06-23
- DOI
- https://doi.org/10.1109/cleo/europe-eqec65582.2025.11111622
- Primary Topic
- Advanced Memory and Neural Computing
- Type
- article
- Field-Weighted Citation Impact
- 0.00