Method for generating instruction sequence and method and device for executing neural network operation

A neural network and instruction sequence technology, applied in the field of artificial intelligence, can solve problems such as large data volume, low hardware efficiency, and insufficient processing power, and achieve the effects of reducing running time, reducing power consumption, and reducing the number of memory accesses

Active Publication Date: 2019-06-21
HORIZON ROBOTICS SHANGHAI ARTIFICIAL INTELLIGENCE TECH CO LTD
View PDF6 Cites 21 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

In the case of limited cache memory capacity, for the application scenario of processing high-resolution images through neural netw...

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
  • Method for generating instruction sequence and method and device for executing neural network operation
  • Method for generating instruction sequence and method and device for executing neural network operation
  • Method for generating instruction sequence and method and device for executing neural network operation

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0024] Hereinafter, an exemplary embodiment according to the present application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments of the present application, and it should be understood that the present application is not limited by the exemplary embodiments described herein.

[0025] Application overview

[0026] With the continuous development of convolutional neural networks (Convoltional Neural Networks, CNN), more application scenarios have been expanded, some of which not only require accuracy, but are also very sensitive to power consumption costs. Especially in some embedded devices, in order to reduce the cost, the capacity of the cache memory (for example, Static Random-Access Memory (SRAM)) will be reduced as much as possible, but in order to ensure the accuracy of the neural network, it is necessary to use a The hi...

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 invention discloses a method for generating an instruction sequence and a method and a device for executing neural network operation. The method for generating the instruction sequence of the neural network comprises the following steps: acquiring attribute data of each layer of the neural network; determining at least one hierarchical group and splitting parameters thereof according to the attribute data of each layer of the neural network, the hierarchical group comprising at least one layer in the neural network, the splitting parameters comprising at least one predetermined dimension and a predetermined number of parts in to which each predetermined dimension is split; and generating an instruction sequence of the neural network according to the attribute data of each layer of theneural network and the splitting parameters of the at least one layer group. According to the embodiment of the invention, all the layers in the neural network are grouped to realize simultaneous splitting calculation of all the layers in the neural network, and under the condition that the capacity of the memory is limited, the efficiency and the processing capability of executing neural networkoperation by hardware can be improved on the premise that the hardware cost is not increased.

Description

Technical field [0001] This application relates to the field of artificial intelligence technology, and in particular to a method for generating an instruction sequence applied to a neural network, a method and a device for performing neural network operations. Background technique [0002] Neural networks have been successfully applied in many fields such as image processing and autonomous driving. In the case of limited cache memory capacity, for application scenarios that process high-resolution images through neural networks, due to the huge amount of data, the hardware is inefficient in performing neural network operations and has insufficient processing capabilities. [0003] Therefore, it is desirable to provide a new method and device applied to neural networks, which can simultaneously improve the efficiency and processing capacity of the hardware to perform neural network operations on the premise of not increasing the hardware cost under the condition of limited memory c...

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
IPC IPC(8): G06N3/063
Inventor 王振江李建军凌坤陈亮黄畅
Owner HORIZON ROBOTICS SHANGHAI ARTIFICIAL INTELLIGENCE TECH CO LTD
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

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

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap