Photonic pattern reconstruction enabled by on-chip online learning and inference

Recent investigations in neuromorphic photonics exploit optical device physics for neuron models, and optical interconnects for distributed, parallel, and analog processing. Integrated solutions enabled by silicon photonics enable high-bandwidth, low-latency and low switching energy, making it a promising candidate for special-purpose artificial intelligence hardware accelerators. Here, we experimentally demonstrate a silicon photonic chip that can perform training and testing of a Hopfield network, i.e. recurrent neural network, via vector dot products. We demonstrate that after online training, our trained Hopfield network can successfully reconstruct corrupted input patterns.

[1]  K. Pagiamtzis,et al.  Content-addressable memory (CAM) circuits and architectures: a tutorial and survey , 2006, IEEE Journal of Solid-State Circuits.

[2]  Paul R. Prucnal,et al.  Silicon Photonic Modulator Neuron , 2018, Physical Review Applied.

[3]  Qing Wu,et al.  Long short-term memory networks in memristor crossbar arrays , 2018, Nature Machine Intelligence.

[4]  L. Chrostowski,et al.  Wavelength tuning and stabilization of microring-based filters using silicon in-resonator photoconductive heaters. , 2015, Optics express.

[5]  J J Hopfield,et al.  Neural networks and physical systems with emergent collective computational abilities. , 1982, Proceedings of the National Academy of Sciences of the United States of America.

[6]  Andrew S. Cassidy,et al.  A million spiking-neuron integrated circuit with a scalable communication network and interface , 2014, Science.

[7]  C. Wright,et al.  Photonics for artificial intelligence and neuromorphic computing , 2020, ArXiv.

[8]  Paul R. Prucnal,et al.  Broadcast and Weight: An Integrated Network For Scalable Photonic Spike Processing , 2014, Journal of Lightwave Technology.

[9]  Bhavin J. Shastri,et al.  Demonstration of scalable microring weight bank control for large-scale photonic integrated circuits , 2020 .

[10]  M. Alava,et al.  Predicting Creep Failure from Cracks in a Heterogeneous Material using Acoustic Emission and Speckle Imaging , 2019, Physical Review Applied.

[11]  Pat Hanrahan,et al.  Understanding the efficiency of GPU algorithms for matrix-matrix multiplication , 2004, Graphics Hardware.

[12]  Corey J. Maley.  Toward Analog Neural Computation , 2017, Minds and Machines.

[13]  C. David Wright,et al.  In-memory computing on a photonic platform , 2018, Science Advances.

[14]  Paul R. Prucnal,et al.  Microring Weight Banks , 2016, IEEE Journal of Selected Topics in Quantum Electronics.

[15]  Philip Y. Ma,et al.  Feedback control for microring weight banks. , 2018, Optics express.

[16]  M. Chagnon,et al.  Design, analysis, and transmission system performance of a 41 GHz silicon photonic modulator. , 2015, Optics express.

[17]  P. Prucnal,et al.  Graphene-based photonic synapse for multi wavelength neural networks , 2020, MRS Advances.

[18]  Dirk Englund,et al.  Deep learning with coherent nanophotonic circuits , 2017, 2017 Fifth Berkeley Symposium on Energy Efficient Electronic Systems & Steep Transistors Workshop (E3S).

[19]  Paul R. Prucnal,et al.  Digital Electronics and Analog Photonics for Convolutional Neural Networks (DEAP-CNNs) , 2019, IEEE Journal of Selected Topics in Quantum Electronics.

[20]  J. Feldmann,et al.  All-optical spiking neurosynaptic networks with self-learning capabilities , 2019, Nature.