Photonic neuromorphic computing has emerged as a promising approach to building a low-latency and energy-efficient non-von Neuman computing system. A photonic spiking neural network (PSNN) exploits brain-like spatiotemporal processing to realize high-performance neuromorphic computing. However, the nonlinear computation of a PSNN remains a significant challenge. Here, we propose and fabricate a photonic spiking neuron chip based on an integrated Fabry–Perot laser with a saturable absorber (FP-SA). The nonlinear neuron-like dynamics including temporal integration, threshold and spike generation, a refractory period, inhibitory behavior and cascadability are experimentally demonstrated, which offers an indispensable fundamental building block to construct the PSNN hardware. Furthermore, we propose time-multiplexed temporal spike encoding to realize a functional PSNN far beyond the hardware integration scale limit. PSNNs with single/cascaded photonic spiking neurons are experimentally demonstrated to realize hardware-algorithm collaborative computing, showing the capability to perform classification tasks with a supervised learning algorithm, which paves the way for a multilayer PSNN that can handle complex tasks.

We proposed and experimentally demonstrated a simple and novel photonic spiking neuron based on a distributed feedback (DFB) laser chip with an intracavity saturable absorber (SA). The DFB laser with an intracavity SA (DFB-SA) contains a gain region and an SA region. The gain region is designed and fabricated by the asymmetric equivalent π-phase shift based on the reconstruction-equivalent-chirp technique. Under properly injected current in the gain region and reversely biased voltage in the SA region, periodic self-pulsation was experimentally observed due to the Q-switching effect. The self-pulsation frequency increases with the increase of the bias current and is within the range of several gigahertz. When the bias current is below the self-pulsation threshold, neuronlike spiking responses appear when external optical stimulus pulses are injected. Experimental results show that the spike threshold, temporal integration, and refractory period can all be observed in the fabricated DFB-SA chip. To numerically verify the experimental findings, a time-dependent coupled-wave equation model was developed, which described the physics processes inside the gain and SA regions. The numerical results agree well with the experimental measurements. We further experimentally demonstrated that the weighted sum output can readily be encoded into the self-pulsation frequency of the DFB-SA neuron. We also benchmarked the handwritten digit classification task with a simple single-layer fully connected neural network. By using the experimentally measured dependence of the self-pulsation frequency on the bias current in the gain region as an activation function, we can achieve a recognition accuracy of 92.2%, which bridges the gap between the continuous valued artificial neural networks and spike-based neuromorphic networks. To the best of our knowledge, this is the first experimental demonstration of a photonic integrated spiking neuron based on a DFB-SA, which shows great potential to realizing large-scale multiwavelength photonic spiking neural network chips.

As Moore’s law has reached its limits, it is becoming increasingly difficult for traditional computing architectures to meet the demands of continued growth in computing power. Photonic neural computing has become a promising approach to overcome the von Neuman bottleneck. However, while photonic neural networks are good at linear computing, it is difficult to achieve nonlinear computing. Here, we propose and experimentally demonstrate a coherent photonic spiking neural network consisting of Mach–Zehnder modulators (MZMs) as the synapse and an integrated quantum-well Fabry–Perot laser with a saturable absorber (FP-SA) as the photonic spiking neuron. Both linear computation and nonlinear computation are realized in the experiment. In such a coherent architecture, two presynaptic signals are modulated and weighted with two intensity modulation MZMs through the same optical carrier. The nonlinear neuron-like dynamics including temporal integration, threshold, and refractory period are successfully demonstrated. Besides, the effects of frequency detuning on the nonlinear neuron-like dynamics are also explored, and the frequency detuning condition is revealed. The proposed hardware architecture plays a foundational role in constructing a large-scale coherent photonic spiking neural network.

No abstract

Spiking neural networks (SNNs) offer powerful computation capability due to its event-driven nature and temporal processing. However, it is still limited to shallow structure and simple tasks due to the training difficulty. In this work, we propose a deep convolutional residual spiking neural network (DCRSNN) for text classification tasks. In the DCRSNN, the feature extraction is achieved via a convolution SNN with residual connection, using the surrogate gradient direct training technique. Classification is performed by a fully-connected network. We also suggest a hybrid photonic DCRSNN, in which photonic SNNs are used for classification with a converted training method. The accuracy of hard and soft reset methods, as well as three different surrogate functions, were evaluated and compared across four different datasets. Results indicated a maximum accuracy of 76.36% for MR, 91.03% for AG News, 88.06% for IMDB and 93.99% for Yelp review polarity. Soft reset methods used in the deep convolutional SNN yielded slightly better accuracy than their hard reset counterparts. We also considered the effects of different pooling methods and observation time windows and found that the convergence accuracy achieved by convolutional SNNs was comparable to that of convolutional neural networks under the same conditions. Moreover, the hybrid photonic DCRSNN also shows comparable testing accuracy. This work provides new insights into extending the SNN applications in the field of text classification and natural language processing, which is interesting for the resources-restrained scenarios.