Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

A method and device for calculating sparse neural network based on systolic array

A systolic array and neural network technology, applied in the field of artificial intelligence, can solve problems such as large resource consumption

Active Publication Date: 2021-04-27
南京风兴科技有限公司
View PDF3 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] This application provides a sparse neural network calculation method based on a systolic array to solve the problems of low efficiency in the encryption and decryption process and large resource consumption in the prior art

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • A method and device for calculating sparse neural network based on systolic array
  • A method and device for calculating sparse neural network based on systolic array
  • A method and device for calculating sparse neural network based on systolic array

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0049] see figure 1 , which is a flow chart of a sparse neural network computing method based on a systolic array architecture of the present application.

[0050] Depend on figure 1 It can be seen that a sparse neural network computing method based on a systolic array architecture provided by an embodiment of the present application includes:

[0051] S100: Obtain a feature map including n weights; the size of the feature map is x*y;

[0052] In this embodiment, the acquired feature map comes from the data input into the computing framework by the previous stage, and the n weights represent the number of sub-feature blocks to be divided into the feature map, so that after the feature map is divided into blocks, each sub-feature block corresponds to a weight. The feature map can usually be represented in the form of x*y, such as 64*8, 128*8 and so on.

[0053] S200: Divide the feature map into n sub-feature blocks equally along the x-axis direction, and the size of each su...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The present application discloses a sparse neural network calculation method based on a systolic array, which includes obtaining a feature map containing n weights; the size of the feature map is x*y; dividing the feature map along the x-axis direction into n sub-feature blocks, the size of each sub-feature block is (x / n)*y; each sub-feature block corresponds to one of the weights; according to the position of the weight in the weight matrix, each sub-feature The block performs calculations to obtain calculation results; and regenerates the weight matrix according to the calculation results and outputs them. In this application, the sparse convolution calculation is realized by means of a systolic array, which makes the data reuse more sufficient, and the convolution is performed in a block manner, which makes the calculation more flexible and more efficient. The weights are encoded and then input to the lower-level equipment. Only Non-zero weights are input to the architecture for calculation, reducing the loss of encoding units.

Description

technical field [0001] The present application relates to the technical field of artificial intelligence, and in particular, to a systolic array-based sparse neural network computing method and device. Background technique [0002] With the continuous development of artificial intelligence (AI), it has evolved from early artificial feature engineering to today's learning from massive data, and can be applied to many fields such as machine vision, speech recognition, and natural language processing. Convolutional Neural Network (CNN) is one of the most representative network structures in deep learning technology, and it is increasingly favored in the field of artificial intelligence, especially in the field of image processing. [0003] As the network coverage becomes wider and more complex, the computing resources for convolution training also grow exponentially, and for neural networks with more layers and nodes, reducing their storage and computing costs becomes critical ...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Patents(China)
IPC IPC(8): G06N3/04
CPCG06N3/045
Inventor 陶为王中风刘文剑谢逍茹
Owner 南京风兴科技有限公司
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com