According to the MIT Technology Review magazine website recently reported that the United States Princeton University research team recently developed the world's first photon neuromorphic chip, and proved its ability to ultra-fast calculation. The chip is expected to open a new photonics industry. The new achievement of Princeton's Tate team at Princeton University is the use of photons to solve the problem of limited speed of neural network circuits. Neural network circuits have set off a storm in computing. The key for scientists to create more powerful neural network circuits is to create circuits that work like neurons, or neuromorphic chips, but the main issue with such circuits is to increase their speed. Photon computing is the "star of tomorrow" in the field of computational science. Photons have more bandwidth than electrons and can process more data quickly. Photonic data processing systems, however, are expensive to manufacture and have therefore not been widely adopted. At the heart of a photon-based neural network developed by the team is an optical device - each of which has neuron-like response characteristics. These nodes take the form of mini circular waveguides that are etched into a silicon base in which light can circulate. When light is input, the output of the laser operating at the threshold is then adjusted, in which small changes in the incident light can have a dramatic effect on the laser's output. The principle of this optical device is that each node in the system uses a certain wavelength of light, a technique known as wavelength-division multiplexing. Light from each node is fed into the laser, and the laser output is fed back to the node, creating a feedback circuit with non-linear features. Research on the extent to which this nonlinear behavior mimics neural behavior shows that its output is mathematically equivalent to a device called Continuous Time Recurrent Neural Network (CTRNN), suggesting that CTRNN programming tools can be applied to more Large silicon photon neural network. The Tate team simulated the mathematical problems of a differential equation with a 49-node silicon photonic neural network and compared it to a common central processing unit. The results show that in this task, the speed of photon neural network has been improved by 3 orders of magnitude. The researchers said it will open a whole new industry in photonics. Tate said: "Silicon photonic neural networks may become the vanguard of a larger, scalable information processing silicon photonics family '."