Deep learning processing device and method supporting encoding and decoding

Abstract

The invention relates to a deep learning processing device and method supporting encoding and decoding, and the device comprises a memory access unit which is used for reading and writing data in a memory; the instruction caching unit is connected to the memory access unit and is used for reading an instruction of a neural network through the memory access unit; the controller unit is connected tothe instruction cache unit; the parameter storage unit is connected to the memory access unit; the parameter decompression unit is connected to the parameter storage unit; and the arithmetic unit isconnected to the parameter storage unit, the parameter decompression unit and the controller unit. Through cooperation of all the units of the device, the compressed parameters can be used for operation, so that the model size of the neural network is effectively reduced, the requirement for the memory is reduced, and the data processing speed of the neural network is effectively increased.

Description

technical field [0001] The present disclosure relates to the field of computer technology, and in particular to a deep learning processing device and method supporting encoding and decoding. Background technique [0002] In recent years, multi-layer artificial neural networks have been widely used in the fields of pattern recognition, image processing, function approximation and optimization calculation. Multi-layer artificial neural network technology has attracted extensive attention from academia and industry due to its high recognition accuracy and good mergeability. However, when it is applied to actual projects, due to the need for The amount of calculation is large, and the model has high memory requirements, so it is difficult to apply it to embedded systems. [0003] In the prior art, general-purpose processors are usually used to process multi-layer artificial neural network operations, training algorithms and compression coding thereof, and support the above-ment...

AI Technical Summary

Problems solved by technology

However, since the GPU is a device specially used to perform graphics and image operations and scientific calculations, there is no support for multi-layer artificial neural networks, so a lot of front-end coding work is required to support multi-layer artificial neural network operations, which brings additional overhead

What's more, the GPU has only a small on-chip cache, and the model data (weights) of the multi-layer artificial neural network need to be repeatedly transported from off-chip, and the GPU cannot compress the model data of the artificial neural network, so it brings huge performance. consumption

Images