2021/07/14 by Matéj Hejda, Hejda, Matěj, Juan Arturo Alanis +14
Computer Science · Engineering · Neuroscience · #Advanced Memory and Neural Computing #Applied Physics (physics.app-ph) #Emerging Technologies (cs.ET) #FOS: Computer and information sciences #FOS: Physical sciences #Neural Networks and Applications #Neural Networks and Reservoir Computing #Neural and Evolutionary Computing (cs.NE) #Neural dynamics and brain function #Optics (physics.optics)
paper · pdf · doi:10.48550/arxiv.2107.06721
openalex publication_date 2021/07/14 · openalex created_date 2022/07/25 · openalex updated_date 2026/07/28
In this work, we introduce an optoelectronic spiking artificial neuron\ncapable of operating at ultrafast rates (\≈ 100 ps/optical spike) and\nwith low energy consumption (< pJ/spike). The proposed system combines an\nexcitable resonant tunnelling diode (RTD) element exhibiting negative\ndifferential conductance, coupled to a nanoscale light source (forming a master\nnode) or a photodetector (forming a receiver node). We study numerically the\nspiking dynamical responses and information propagation functionality of an\ninterconnected master-receiver RTD node system. Using the key functionality of\npulse thresholding and integration, we utilize a single node to classify\nsequential pulse patterns and perform convolutional functionality for image\nfeature (edge) recognition. We also demonstrate an optically-interconnected\nspiking neural network model for processing of spatiotemporal data at over 10\nGbps with high inference accuracy. Finally, we demonstrate an off-chip\nsupervised learning approach utilizing spike-timing dependent plasticity for\nthe RTD-enabled photonic spiking neural network. These results demonstrate the\npotential and viability of RTD spiking nodes for low footprint, low energy,\nhigh-speed optoelectronic realization of neuromorphic hardware.\n