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

A method for realizing state quantization network in cross-array neural morphology hardware

A cross-array, neuromorphic technology, applied in the field of neural networks, to achieve the effect of reducing scale

Active Publication Date: 2019-01-08
UNIV OF ELECTRONICS SCI & TECH OF CHINA
View PDF7 Cites 11 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0007] In the currently implemented cross-array neuromorphic hardware, parameters such as synaptic weights and neuron parameters such as thresholds, leakage constants, set voltages, refractory periods, and synaptic delays require a lot of system storage resources. With the rapid expansion of circuit scale, this will inevitably become a major bottleneck of neuromorphic hardware under the condition that storage resources are relatively scarce today.

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 for realizing state quantization network in cross-array neural morphology hardware
  • A method for realizing state quantization network in cross-array neural morphology hardware
  • A method for realizing state quantization network in cross-array neural morphology hardware

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0034] In order to make the object, technical solution and advantages of the present invention clearer, the implementation manner of the present invention will be further described in detail below in conjunction with the accompanying drawings.

[0035] refer to figure 1 , the embodiment of the present invention provides a method for implementing a state quantization network in a cross-array neuromorphic hardware, comprising the following steps:

[0036] S1: Select parameters and quantize them. Parameter quantization can be performed after the neural network training is completed, or during neural network training.

[0037] A: Quantize after the neural network training is complete

[0038] The artificial neural network (including MLP, CNN, RNN, LSTM, etc.) is trained to obtain parameters under specific tasks and conditions, (including weights, thresholds, leakage constants, set voltage values, refractory period duration, synaptic delay duration, etc.);

[0039] The artificia...

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 belongs to the technical field of neural networks and relates to a method for realizing state quantization network in cross array neural morphology hardware. The method of the inventionincludes: after quantifying the parameters of an artificial neural network (weight, threshold, leakage constant, set voltage, refractory period and synaptic delay), the quantized parameters are mappedto the cross-array neural morphology hardware, and then the input data after preprocessing are sent to the cross-array neural morphology hardware to realize the state quantization network. Through state quantization, the requirements of cross-array neural morphology hardware for memory cell size, memory level, reliability and so on are effectively reduced.

Description

technical field [0001] The invention belongs to the technical field of neural networks, and relates to a method for realizing a state quantization network in cross-array neuromorphic hardware. Background technique [0002] Neuromorphic computing is used to refer to brain-derived computers, devices, and models in contrast to the prevalent von Neumann computer architecture. This biomimetic approach creates highly connected synthetic neurons and synapses that can be used for theoretical modeling in neuroscience to solve machine learning problems. [0003] Neuromorphic circuit is one of the physical realizations of the neural network model. It abstracts and simulates the biological nervous system at a high level by means of hardware, in order to achieve low power consumption, High adaptability and other characteristics. [0004] Crossbars use memristors for data storage and parallel computing and as an important component architecture of neural network nodes is to use crossbar...

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 Applications(China)
IPC IPC(8): G06N3/06G06N3/063
CPCG06N3/061G06N3/063
Inventor 胡绍刚罗鑫乔冠超刘益安张成明刘洋于奇
Owner UNIV OF ELECTRONICS SCI & TECH OF CHINA
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