15 April 2025 in Light: Science & Applications
Multi-wavelength optical information processing systems are commonly utilized in optical neural networks and broadband signal processing. However, their effectiveness is often compromised by frequency-selective responses caused by fabrication, transmission, and environmental factors. To mitigate these issues, this study introduces a deep reinforcement learning calibration (DRC) method inspired by the deep deterministic policy gradient training strategy. This method continuously and autonomously learns from the system, effectively accumulating experiential knowledge for calibration strategies and demonstrating superior adaptability compared to traditional methods. In systems based on dispersion compensating fiber, micro-ring resonator array, and Mach-Zehnder interferometer array that use multi-wavelength optical carriers as the light source, the DRC method enables the completion of the corresponding signal processing functions within 21 iterations. This method provides efficient and accurate control, making it suitable for applications such as optical convolution computation acceleration, microwave photonic signal processing, and optical network routing.
16 August 2021 in Laser & Photonics Reviews
1 November 2024 in Photonics Research
In recent years, integrated optical processing units (IOPUs) have demonstrated advantages in energy efficiency and computational speed for neural network inference applications. However, limited by optical integration technology, the practicality and versatility of IOPU face serious challenges. In this work, a scalable parallel photonic processing unit (SPPU) for various neural network accelerations based on high-speed phase modulation is proposed and implemented on a silicon-on-insulator platform, which supports parallel processing and can switch between multiple computational paradigms simply and without latency to infer different neural network structures, enabling to maximize the utility of on-chip components. The SPPU adopts a scalable and process-friendly architecture design, with a preeminent photonic-core energy efficiency of 0.83 TOPS/W, two to ten times higher than existing integrated solutions. In the proof-of-concept experiment, a convolutional neural network (CNN), a residual CNN, and a recurrent neural network (RNN) are all implemented on our photonic processor to handle multiple tasks of handwritten digit classification, signal modulation format recognition, and review emotion recognition. The SPPU achieves multi-task parallel processing capability, serving as a promising and attractive research route to maximize the utility of on-chip components under the constraints of integrated technology, which helps to make IOPU more practical and universal.
13 August 2024 in Optics Express
In the rapidly evolving field of artificial intelligence, integrated photonic computing has emerged as a promising solution to address the growing demand for high-performance computing with ultrafast speed and reduced power consumption. This study presents what we believe is a novel photonic tensor processing core (PTPC) on a chip utilizing wavelength division multiplexing technology to perform parallel multiple vector-matrix multiplications concurrently, allowing for reconfigurable computing dimensions without changing the hardware scale. Specifically, this architecture significantly enhances the number of operations in convolutional neural networks, making it superior to other photonic computing systems. Experimental evaluations demonstrate the high-speed performance of the PTPC, achieving an impressive total computing speed of 0.252 TOPS and a computing speed per unit as high as 0.06 TOPS /unit in a compact hardware scale. Additionally, proof-of-concept application experiments are conducted on benchmark datasets, including the Modified National Institute of Standards and Technology (MNIST), Google Quickdraw, and CIFAR-10, with high accuracies of 97.86%, 93.51%, and 70.22%, respectively, in image recognition and classification tasks. By enabling parallel operations in PTPC on a chip, this study opens new avenues for exploration and innovation at the intersection of silicon photonics, scalable computation, and artificial intelligence, shaping the future landscape of computing technologies.
1 November 2021 in Optics Letters
We propose an object recognition architecture relying on a neural network algorithm in optical sensors. Precisely, by applying the high-speed and low-power Fourier transform operation in the optical domain, we can transfer the high-cost part of the traditional convolutional neural network algorithm to the sensor side to achieve faster computing speed. An optical neuron unit (ONU) consisting of transition metal sulfide (TMD) material is fabricated for a vivid validation of this architecture. Using the embedded gate pair structure inside our ONU, TMD materials can be electrically doped at different levels, forming an in-plane PN junction, which allows for effective manipulation of light response to imitate biological nerve synapses. The results demonstrate that our ONU could reach the ability of optic neurons, providing experimental support for future in-sensor computing architecture.