Processing system and method for three-value weight convolution network

Abstract

The invention provides a processing system for three-value weight convolution network. The system comprises at least one storage unit which is used for storing data and an instruction; at least one control unit which is used for obtaining the instruction stored in the storage unit and transmitting a control signal; and at least one calculating unit which is used for obtaining a node value of one layer in a convolution neural network from the storage unit and the corresponding three-value weight data, and obtaining the node value of a next layer through the execution of addition and subtraction operations. The system reduces the data bit width in the calculation process of the convolution neural network, improves the convolution calculation speed, and reduces the storage capacity and work energy consumption.

Description

technical field [0001] The invention relates to the field of computer technology, in particular to a processing system applied to a ternary weight convolution network. Background technique [0002] Deep learning technology has developed rapidly in recent years. Deep neural networks, especially convolutional neural networks, have made great achievements in image recognition, speech recognition, natural language understanding, weather prediction, gene expression, content recommendation and intelligent robots. Wide range of applications. The deep network structure obtained through deep learning is an operational model, which contains a large number of data nodes, each data node is connected to other data nodes, and the connection relationship between each node is represented by weight. With the continuous improvement of the complexity of the neural network, the neural network technology has many problems in the actual application process, such as occupying a lot of resources, ...

AI Technical Summary

Problems solved by technology

[0004] However, most of the current deep learning applications are implemented using central processing units and graphics processing units. These technologies are not energy efficient, and there are serious energy efficiency problems and computing speed bottlenecks when applied in embedded devices or low-overhead data centers. , it is difficult to meet the performance requirements of the application, so it is difficult to apply it to small and light-weight devices such as mobile phones and embedded electronic devices

Images