A review: Photonics devices, architectures, and algorithms for optical neural computing

Abstract

The explosive growth of data and information has motivated various emerging non-von Neumann computational approaches in the More-than-Moore era. Photonics neuromorphic computing has attracted lots of attention due to the fascinating advantages such as high speed, wide bandwidth, and massive parallelism. Here, we offer a review on the optical neural computing in our research groups at the device and system levels. The photonics neuron and photonics synapse plasticity are presented. In addition, we introduce several optical neural computing architectures and algorithms including photonic spiking neural network, photonic convolutional neural network, photonic matrix computation, photonic reservoir computing, and photonic reinforcement learning. Finally, we summarize the major challenges faced by photonic neuromorphic computing, and propose promising solutions and perspectives.

Recent progress of integrated circuits and optoelectronic chips

“…There are various prototypical neuromorphic processers, including SpiNNaker, Neurogrid, TrueNorth, Loihi, and Tianjic [173]. As an alternative, photonics neuromorphic, which is an emerging field that combines the photonics and neuroscience, has attracted lots of attention due to the fascinating advantages such as high speed, wide bandwidth, and massive parallelism [174][175][176][177][178].…”

“…From the device perspective, it has been proved that various optical devices can emulate the neuron-like [179][180][181] and synapse-like functions [182,183]. There have been numerous attempts to implement photonics spiking neurons like graphene excitable lasers, distributed feedback lasers, vertical-cavity surface-emitting lasers, or micropillars [174,178]. The spike temporal encoding, excitatory and inhibitory response, refractory period, and spike-timing dependent plasticity (STDP) have been demonstrated numerically and experimentally.…”

“…Hao Y, et al Sci China Inf Sci Figure 17 (Color online) Timeline of advances in photonics neuromorphic[174,178]. VCSEL: vertical-cavity surface-emitting lasers; VCSEL-SA: vertical-cavity surface-emitting lasers with embedded saturable; SOA: semiconductor optical amplifier; EAM: electro-absorption modulator; MRR: microring resonator; MZI: Mach-Zehnder interferometer; VCSOA: vertical-cavity semiconductor optical amplifier; DFB: distributed feedback laser; WTA: winner-take-all.…”

Implementation of Pruned Backpropagation Neural Network Based on Photonic Integrated Circuits

“…Photonic integrated circuits can easily realize matrix multiplication with the coherence and superposition of linear optics [9]. The advantages of photonics have promoted extensive demonstrations of various high-performance optical functionalities, especially in photon computing [10][11][12][13][14][15]. The programmable Mach Zehnder interferometers realize optical interference units for optical neural networks [11,16], and MRRs are used for optical matrixvector multipliers to implement neuromorphic photonic networks [17][18][19].…”

Optical Vortices Sharp Focusing by Silicon Ring Gratings Using High-Performance Computer Systems

“…It is also possible to obtain a high integration density due to the large refractive index of silicon [1]. Photonics in general and silicon photonics in particular are becoming a promising alternative for the implementation of new accelerators for deep learning that can use light, both for communication and for computing, thanks to such advantages as high speed, wide bandwidth and parallelism [6,7].…”