Circuit structure and its driving method, neural network

A circuit structure and neural network technology, applied in the field of neural networks, can solve the problems of high energy consumption of electronic devices, inability to handle complex tasks, and limiting the development of brain-like computing systems.

Active Publication Date: 2020-05-05
TSINGHUA UNIV
View PDF5 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0003] At present, in the brain-inspired computing system, neuron-like functions need to be realized through circuit design. Due to the limitation of hardware resources, neuron-like electronic devices have defects such as high energy consumption and inability to handle complex tasks, which further limit the brain-like computing system. development of

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
  • Circuit structure and its driving method, neural network
  • Circuit structure and its driving method, neural network
  • Circuit structure and its driving method, neural network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0079] Figure 2a A schematic diagram of a circuit structure provided by this embodiment is shown.

[0080] For example, if Figure 2aAs shown, the circuit structure 100 includes: at least one circuit unit 101, each circuit unit 101 includes a first group of resistive switching devices 1 and a second group of resistive switching devices 2, and the first group of resistive switching devices 1 includes a resistance variable device 10 , the second group of resistive switching devices 2 includes a sudden resistance device 20, the first group of resistive switching devices 1 and the second group of resistive switching devices 2 are connected in series, and when no voltage is applied, the resistance of the first group of resistive switching devices 1 The value is greater than the resistance value of the second group of resistive switching devices 2 .

[0081] In the circuit structure, the resistance gradient device and the resistance mutation device are connected in series to form...

Embodiment 2

[0123] This embodiment provides a neural network, Figure 6a A block diagram of the neural network provided in this embodiment; Figure 6b A schematic structural diagram of the neural network provided in this embodiment; Figure 6c-6e Schematic diagrams of the structures of different neuron units provided for this example.

[0124] For example, if Figure 6a As shown, the neural network includes: a neuron unit 200 . The neuron unit 200 includes the circuit structure 100 described in any one of the above embodiments.

[0125] For example, the neuron unit 200 further includes at least one signal line. One end of the circuit unit 101 is connected to the signal line, and the other end of the circuit unit 101 may be connected to the ground line. For example, if Figure 6c As shown, the neuron unit 200 also includes a first signal line 50, the first signal line 50 is electrically connected to the first group of resistive switching devices 1 or the second group of resistive swi...

Embodiment 3

[0136] This embodiment provides a driving method for the circuit structure described in any one of the first embodiment, Figure 7 The working sequence diagram of the driving method provided in this embodiment.

[0137] For example, the driving method includes the following operations: For example, using a driver, applying a voltage signal to the circuit unit through a signal line to change the resistance value of the first group of resistive switching devices, thereby performing signal integration; when the resistance value of the second group of resistive switching devices When a mutation occurs, a signal is emitted.

[0138] This operation method reduces the resistance value of the first group of resistive switching devices by applying a voltage signal to the circuit unit, so that the current in the circuit unit increases slowly, thereby realizing the integral function; when the resistance value of the second group of resistive switching devices suddenly decreases , the cu...

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

A circuit structure, its driving method, and neural network. The circuit structure includes: at least one circuit unit, each circuit unit includes a first group of resistive switching devices and a second group of resistive switching devices, the first group of resistive switching devices includes resistance gradient devices, and the second group of resistive switching devices includes resistance switching devices The abrupt change device, the first group of resistive switching devices and the second group of resistive switching devices are connected in series, and the resistance value of the first group of resistive switching devices is greater than the resistance value of the second group of resistive switching devices when no voltage is applied. In the circuit structure, the resistance gradient device and the resistance mutation device are connected in series to form a neuron-like structure, so as to realize the function of simulating human brain neuron. The resistance value of the resistance gradient device changes slowly under the applied voltage, which can be used to simulate the behavior of the S-shaped growth curve of biological dendrites. The circuit structure has the advantages of simple structure, low power consumption, small area, complex functions and easy compatibility with standard CMOS technology.

Description

technical field [0001] Embodiments of the present disclosure relate to a circuit structure, a driving method thereof, and a neural network. Background technique [0002] With the rapid development of information technology, the improvement of operating speed and the reduction of energy consumption have become the bottleneck of the development of the traditional von Neumann computing architecture. Features such as adaptation are considered to be the breakthrough direction for high-performance computers. In the human brain, neurons and synapses are the most important and most numerous neural units. Therefore, building a brain-like computing system requires a large number of neuron-like and synapse-like electronic devices. In addition to the functions of simulating neurons and synapses, these electronic devices also need to have small area, low power consumption, and high speed. performance. [0003] At present, in the brain-inspired computing system, neuron-like functions n...

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): G06N3/063
CPCG06N3/065G06N3/049G06N3/04G06N3/063
Inventor 李辛毅吴华强宋森张清天高滨钱鹤
Owner TSINGHUA UNIV
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