Three-dimensional oscillator structure

Patent metadata

Field	Value
Publication number	US-11250317-B2
Application number	US-201816144952-A
Country	US
Kind code	B2
Filing date	Sep 27, 2018
Priority date	Sep 27, 2018
Publication date	Feb 15, 2022
Grant date	Feb 15, 2022

Abstract

Embodiments may relate to a structure to be used in a neural network. A first column and a second column, both of which are to couple with a substrate. A capacitor structure may be electrically coupled with the first column. An insulator-metal transition (IMT) structure may be coupled with the first column such that the capacitor structure is electrically positioned between the IMT structure and the first column. A resistor structure may further be electrically coupled with the IMT structure and the second column such that the resistor structure is electrically positioned between the second column and the IMT structure. Other embodiments may be described or claimed.

First claim

The invention claimed is:

1. A structure to be used in a neural network, wherein the structure comprises:
a first column to couple with a face of a substrate, wherein the face of the substrate defines a lateral plane parallel to the face of the substrate and a vertical axis perpendicular to the lateral plane, wherein the first column is parallel to the vertical axis;
a second column to couple with the face of the substrate, wherein the second column is parallel to the vertical axis;
a capacitor structure electrically coupled with the first column;
an insulator-metal transition (IMT) structure electrically coupled with the first column, wherein the capacitor structure is electrically positioned between the IMT structure and the first column; and
a resistor structure electrically coupled with the IMT structure and the second column, wherein the resistor structure is electrically positioned between the second column and the IMT structure.

2. The structure of claim 1 , further comprising a voltage line electrically coupled with the IMT structure, wherein the IMT structure is electrically positioned between the voltage line and the capacitor structure.

3. The structure of claim 1 , wherein the first column is to couple with averaging logic of the neural network.

4. The structure of claim 1 , wherein the second column is to couple with ground of the substrate.

5. The structure of claim 1 , wherein the capacitor structure includes a first electrode, a second electrode, and a dielectric material between the first and second electrodes.

6. The structure of claim 1 , wherein the IMT structure includes a first electrode, a second electrode, and an oxide material positioned between the first electrode and the second electrode.

7. The structure of claim 1 , wherein the resistor structure includes a first electrode and a second electrode and a variable resistor positioned between the first electrode and the second electrode.

8. The structure of claim 7 , wherein the variable resistor is a resistive random-access memory (RRAM), a conductive bridge random-access memory (CBRAM), or a phase change material (PCM).

9. The structure of claim 7 , wherein the variable resistor is a filamentary oxygen vacancy resistive switching material, an interfacial resistive switching material, or a conductive bridge filament material.

10. The structure of claim 7 , further comprising a wordline coupled with the variable resistor, wherein the wordline is to deliver a signal related to a desired resistance setting of the variable resistor.

11. The structure of claim 7 , wherein the variable resistor is a floating gate transistor.

12. The structure of claim 1 , wherein the resistor structure includes a wire.

13. A structure that is a portion of a neural network, wherein the structure comprises:
a substrate with a face, wherein the face defines a lateral plane parallel to the face of the substrate and a vertical axis perpendicular to the lateral plane;
a first plurality of columns, wherein respective ones of the first plurality of columns are parallel to the vertical axis and are electrically coupled with averaging logic of the neural network;
a second plurality of columns, wherein respective ones of the second plurality of columns are parallel to the vertical axis and are electrically coupled with a ground of the substrate; and
an oxide oscillator electrically coupled with and electrically positioned between a first column of the first plurality of columns and a second column of the second plurality of columns, wherein the oxide oscillator includes:
a capacitor structure electrically coupled with the first column;
an insulator-metal transition (IMT) structure electrically coupled with the first column, wherein the capacitor structure is electrically positioned between the IMT structure and the first column; and
a resistor structure electrically coupled with the IMT structure and the second column, wherein the resistor structure is electrically positioned between the second column and the IMT structure.

14. The structure of claim 13 , wherein the oxide oscillator is a first oxide oscillator, and further comprising a second oxide oscillator electrically coupled with and electrically positioned between the first column and a third column of the second plurality of columns.

15. The structure of claim 14 , further comprising a first voltage line electrically coupled with the IMT structure of the first oxide oscillator and a second voltage line electrically coupled with the IMT structure of the second oxide oscillator.

16. The structure of claim 14 , wherein the first oxide oscillator and the second oxide oscillator are coupled with one another by the first column.

17. A method of forming a structure for use in a neural network, the method comprising:
coupling a first column with a substrate, wherein the first column is parallel to a vertical axis of the structure, and wherein the vertical axis is perpendicular to a lateral axis of the structure;
coupling a second column to the substrate, wherein the second column is parallel to the first column;
coupling a first oscillator to the first column and the second column, wherein the first oscillator is at a first location along the vertical axis; and
coupling a second oscillator to the first column and the second column, wherein the second oscillator is at a second location along the vertical axis.

18. The method of claim 17 , wherein the first column is coupled with an averaging logic of the neural network, and the second column is coupled with a ground of the substrate.

19. The method of claim 17 , wherein the first oscillator includes:
a capacitor structure electrically coupled with the first column;
an insulator-metal transition (IMT) structure electrically coupled with the first column, wherein the capacitor structure is electrically positioned between the IMT structure and the first column, and wherein the IMT structure is further electrically coupled with a voltage rail such that the IMT structure is electrically positioned between the voltage rail and the capacitor structure; and
a resistor structure electrically coupled with the IMT structure and the second column, wherein the resistor structure is electrically positioned between the second column and the IMT structure.

20. The method of claim 17 , further comprising:
coupling a third column with the substrate; and
coupling a third oscillator to the first column and the third column, wherein the third oscillator is as the first location along the vertical axis.

Assignees

• Intel Corp

Inventors

• Nikonov Dmitri E
• Karpov Elijah V
• Young Ian A

Classifications

• H03B5/20 Primary

  with frequency-determining element comprising resistance and either capacitance or inductance, e.g. phase-shift oscillator · CPC title

• G06N3/065 Primary

  Analogue means · CPC title

• G06N3/049

  Temporal neural networks, e.g. delay elements, oscillating neurons or pulsed inputs · CPC title

• H03K4/787

  using as active elements semiconductor devices with two electrodes and exhibiting a negative resistance characteristic · CPC title

• H03B1/02

  Structural details of power oscillators, e.g. for heating {(construction of transmitters H04B; features of generators for heating by electromagnetic fields H05B6/00)} · CPC title