IP Library Granted Patent US 10,867,237
Granted Patent B2
US 10,867,237 · App. 15/307,269 · Granted Dec 15, 2020

Single-component artificial neuron based on Mott insulators, network of artificial neurons and corresponding manufacturing method

Inventors: Laurent Cario (Nantes, FR); Benoit Corraze (Carquefou, FR); Pablo Stoliar (Ibaraki, JP); Julien Tranchant (Reze, FR); Etienne Janod (La Chapelle sur Erdre, FR); Marie-Paule Besland (Orvault, FR); Marcelo Rozenberg (Paris, FR)
Assignees: CENTRE NATIONAL DE LA RECHERCHE SCIENTIFIQUE—CNRS; UNIVERSITE DE NANTES; UNIVERSITE PARIS-SUD XI
G06N3/063G06N3/049G06N3/0635G11C11/54G11C13/0002H01L45/1253H01L45/14H01L45/16
View Patent ↗
Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US 10,867,237
App. No.
15/307,269
Granted
Dec 15, 2020
Kind
B2
Abstract

An artificial neuron includes a single-component electric dipole including a single material which belongs to the class of Mott insulators and is connected to first and second electric electrodes.

Claims (32)

1. An artificial neuron implementing functions of integration, leakage and firing, wherein the artificial neuron consists of:

a single-component electrical dipole comprising a material belonging to the family of Mott insulators connected to first and second electrical electrodes, the artificial neuron being a Leaky Integrate-and-Fire (LIF) reference model artificial neuron consisting of the single-component electrical dipole.

2. The artificial neuron according to claim 1 , wherein said Mott insulator material comprises:

a compound of formula AM 4 Q 8 , with A comprising at least one of the following elements: Ga, Ge, Zn; M comprising at least one of the following elements: V, Nb, Ta, Mo: and Q comprising at least one of the following elements: S, Se; or

an inorganic compound of formula (V 1-x M x ) 2 O 3 , with 0≤x≤1, M comprising at least one of the following elements: Ti, Cr, Fe, Al, or Ga; or

an inorganic compound of NiS 2-x Se x , with 0≤x≤1; or

a compound of formula VO 2 ; or

an organic Mott insulator compound.

3. The artificial neuron according to claim 1 , wherein the first and second electrical electrodes are each constituted by an electrically conductive material comprising:

one of the following elements: platinum (Pt), gold (Au), molybdenum (Mo), graphite (C), aluminum (Al), copper (Cu), doped silicon (Si); or

one of the following alloys: brass (Cu—Zn), steel (Fe—C), bronze (Cu—Sn); or

one of the following transition metal compounds: TiN, TaN, RuO 2 , SrRuO 3 , CuS 2 .

4. The artificial neuron according to claim 1 , wherein said Mott insulator material takes the form of:

a thin layer; or

a block of crystal; or

a nanotube; or

a nanowire.

5. A network of neurons comprising a plurality of artificial neurons, wherein at least one artificial neuron is according to claim 1 .

6. A neuromorphic electronic circuit comprising a plurality of artificial neurons, wherein at least one artificial neuron is according to claim 1 .

7. A method for manufacturing an artificial neuron implementing the functions of integration, leakage and firing, wherein the method comprises the following acts:

obtaining a material belonging to the family of Mott insulators;

obtaining a Leaky Integrate-and-Fire (LIF) reference model artificial neuron consisting of

a single-component electrical dipole by deposition of a layer of conductive material:

at a first extremity of said Mott insulator material to form a first electrical electrode, and

at a second extremity of said Mott insulator material to form a second electrical electrode.

8. The method for manufacturing according to claim 7 , wherein said act of obtaining a material is performed by cutting out a block of Mott insulator crystal, and wherein said act of depositing a layer of conductive material is performed as a function of said cut-out block of crystal.

9. The method for manufacturing according to claim 7 , wherein said act of obtaining a material is performed by depositing, on a substrate wafer, a thin layer of a Mott insulator material,

and wherein said act of depositing a layer of conductive material is performed as a function of said deposited thin layer.

10. The method for manufacturing according to claim 7 , wherein said act of obtaining a material is performed by depositing, on a substrate wafer, a nanotube based on a Mott insulator material,

and wherein said act of depositing a layer of conductive material is performed as a function of said deposited nanotube.

11. The method for manufacturing according to claim 7 , wherein said act of obtaining a material is performed by depositing, on a substrate wafer, a nanowire based on a Mott insulator material,

and wherein said act of depositing a layer of conductive material is performed as a function of said deposited nanowire.

Assignments (3)
MERGER Recorded Sep 7, 2023
From: UNIVERSITE DE NANTES
To: NANTES UNIVERSITE
Reel/Frame 064824/0596 →
MERGER Recorded Aug 22, 2022
From: UNIVERSITE DE PARIS 11 - PARIS SUD (ALSO NAMED UNIVERSITÉ PARIS-SUD)
To: UNIVERSITE PARIS-SACLAY
Reel/Frame 060856/0245 →
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Sep 12, 2017
From: CARIO, LAURENT; CORRAZE, BENOIT; STOLIAR, PABLO; TRANCHANT, JULIEN; JANOD, ETIENNE; BESLAND, MARIE-PAULE; ROZENBERG, MARCELO
To: CENTRE NATIONAL DE LA RECHERCHE SCIENTIFIQUE - CNRS; UNIVERSITE DE NANTES; UNIVERSITE PARIS-SUD XI
Reel/Frame 043562/0150 →
Priority Claims (1)
FR 14 53834 · Apr 28, 2014 · national
Continuity (1)
Related Publication 20170124449A1 · May 4, 2017