Scale-extensible convolutional neural network acceleration system and method

Abstract

A scale-extensible convolutional neural network acceleration system, and the system comprises a processor and at least one convolutional acceleration core, wherein each convolutional acceleration coremainly comprises a computing array, a controller and an on-chip cache; when the scale is extended, only the number of the convolutional acceleration cores needs to be increased, programs running on the processor need to be modified, and other hardware modules do not need to be changed. Namely, a plurality of convolution acceleration kernels can be added to improve the scale and computing performance of the system. The invention further provides a method based on the scale-extensible convolutional neural network acceleration system, extra overhead caused by scale extension can be reduced to agreat extent, and therefore the scale-extensible convolutional neural network acceleration system can be deployed on different hardware platforms. Meanwhile, the software and hardware cooperation modeis good in universality, and different convolutional neural networks can be supported. Compared with other circuits, the method has universality and expandability.

Description

technical field [0001] The invention belongs to the technical field of artificial intelligence, relates to the improvement of neural network computing performance, and in particular to a scalable convolutional neural network acceleration system and method. Background technique [0002] In recent years, convolutional neural networks have been widely deployed in IoT smart terminals, autonomous driving, and data centers due to their superior performance. Relying on a large amount of training data, network algorithms can be applied to many fields such as image classification and detection, human-computer games, and natural language processing. [0003] However, as the convolutional neural network structure becomes more complex and its depth continues to increase, the amount of network calculations also increases, resulting in problems such as low computing efficiency or high power consumption on general-purpose hardware platforms. Computational neural networks. The design of c...

AI Technical Summary

Problems solved by technology

If you need to expand, you need to modify and design the corresponding system, so it brings a lot of overhead and extra design time

And if it is a separate convolution accelerator, it can only execute a specific convolutional neural network, and does not support part of the calculation in the updated network, which is less flexible.

Images