# US 2024/0201950 A1 · Normalization in analog memory-based neural network

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2024/0201950 A1
- Document ID: US20240201950A1
- Status: Published application
- Kind: A1
- Application number: 18/081,859
- Filed: 2022-12-15
- Published: 2024-06-20
- Earliest claimed: 2022-12-15
- Assignees: International Business Machines Corporation
- Applicants: International Business Machines Corporation
- Inventors: Stefano Ambrogio; Andrea Fasoli
- Web page: https://patentgrep.com/patent/US20240201950A1

## CPC

- G06F 7/5443: {Sum of products (for applications thereof, see the relevant places, e.g. G06F17/10, H03H17/00)}
- G06N 3/065: Analogue means
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS

## Abstract

A scale factor can be determined and applied to input values of an analog neural network implemented by a crossbar array of non-volatile memory devices. An inverse of the scale factor can be applied to synaptic weight values stored by the crossbar array of non-volatile memory devices. Operations by the crossbar array can be performed using the scaled input values and the synaptic weight values. The scale factor can be determined for each row of the crossbar array. The scale factor can be determined using a combination of the input values and the synaptic weight values.

## Claims

### Claim 1

A method comprising:
determining a scale factor;
applying the scale factor to input values of an analog neural network implemented by a crossbar array of non-volatile memory devices; and
applying an inverse of the scale factor to synaptic weight values stored by the crossbar array of non-volatile memory devices.

### Claim 2 (depends on 1)

The method of claim 1, wherein the scale factor is determined for each row of the crossbar array.

### Claim 3 (depends on 2)

The method of claim 2, wherein the scale factor is determined using a maximum of absolute values of synaptic weight values stored on a row of the crossbar array.

### Claim 4 (depends on 2)

The method of claim 2, wherein the scale factor is determined using a maximum of absolute values of input values being applied to a row of the crossbar array.

### Claim 5 (depends on 2)

The method of claim 2, wherein the scale factor is determined using a combination of a maximum of absolute values of synaptic weight values stored on a row of the crossbar array and a maximum of absolute values of input values being applied to the row of the crossbar array.

### Claim 6 (depends on 2)

The method of claim 2, wherein the scale factor is determined using a sum of absolute values of synaptic weight values stored on a row of the crossbar array.

### Claim 7 (depends on 2)

The method of claim 2, wherein the scale factor is determined using a sum of absolute values of input values being applied to a row of the crossbar array.

### Claim 8 (depends on 2)

The method of claim 2, wherein the scale factor is determined using a combination of a sum of absolute values of synaptic weight values stored on a row of the crossbar array and a sum of absolute values of input values being applied to the row of the crossbar array.

### Claim 9

An apparatus comprising:
a crossbar array of non-volatile memory devices configured to perform multiply and accumulate operations based on synaptic weight values of a neural network stored on the non-volatile memory devices and input values received via rows of input lines coupled to the rows of the crossbar array; and
a processor configured to apply a scale factor to the input values and to apply an inverse of the scale factor to the synaptic weight values stored by the crossbar array of non-volatile memory devices.

### Claim 10 (depends on 9)

The apparatus of claim 9, wherein the processor is configured to determine the scale factor for each row of the crossbar array.

### Claim 11 (depends on 10)

The apparatus of claim 10, wherein the scale factor is determined using a maximum of absolute values of synaptic weight values stored on a row of the crossbar array.

### Claim 12 (depends on 10)

The apparatus of claim 10, wherein the scale factor is determined using a maximum of absolute values of input values being applied to a row of the crossbar array.

### Claim 13 (depends on 10)

The apparatus of claim 10, wherein the scale factor is determined using a combination of a maximum of absolute values of synaptic weight values stored on a row of the crossbar array and a maximum of absolute values of input values being applied to the row of the crossbar array.

### Claim 14 (depends on 10)

The apparatus of claim 10, wherein the scale factor is determined using a sum of absolute values of synaptic weight values stored on a row of the crossbar array.

### Claim 15 (depends on 10)

The apparatus of claim 10, wherein the scale factor is determined using a sum of absolute values of input values being applied to a row of the crossbar array.

### Claim 16 (depends on 10)

The apparatus of claim 10, wherein the scale factor is determined using a combination of a sum of absolute values of synaptic weight values stored on a row of the crossbar array and a sum of absolute values of input values being applied to the row of the crossbar array.

### Claim 17

A computer program product comprising a computer readable storage medium having program instructions embodied therewith, the program instructions readable by a device to cause the device to:
determine a scale factor;
apply the scale factor to input values of an analog neural network implemented by a crossbar array of non-volatile memory devices; and
apply an inverse of the scale factor to weight values stored by the crossbar array of non-volatile memory devices.

### Claim 18 (depends on 17)

The computer program product of claim 17, the scale factor is determined for each row of the crossbar array.

### Claim 19 (depends on 18)

The computer program product of claim 18, wherein the scale factor is determined using a combination of a maximum of absolute values of synaptic weight values stored on a row of the crossbar array and a maximum of absolute values of input values being applied to the row of the crossbar array

### Claim 20 (depends on 18)

The computer program product of claim 18, wherein the scale factor is determined using a combination of a sum of absolute values of synaptic weight values stored on a row of the crossbar array and a sum of absolute values of input values being applied to the row of the crossbar array.
