IP Library Granted Patent US 12670376
Granted Patent B2
US 12670376 · App. 17/893,075 · Granted Jun 30, 2026

Testing circuitry and methods for analog neural memory in artificial neural network

Inventors: Hieu Van Tran (San Jose, CA); Thuan Vu (San Jose, CA); Stephen Trinh (San Jose, CA); Stanley Hong (San Jose, CA); Anh Ly (San Jose, CA); Steven Lemke (Boulder Creek, CA); Nha Nguyen (San Jose, CA); Vipin Tiwari (Dublin, CA); Nhan Do (Saratoga, CA)
Assignee: Silicon Storage Technology, Inc.
G06N3/065G06N3/0464G11C16/0425G11C16/10G11C16/3459G11C16/08
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 12670376
App. No.
17/893,075
Granted
Jun 30, 2026
Kind
B2
Abstract

In one example, a method comprises programming a plurality of analog neural non-volatile memory cells in an array of analog neural non-volatile memory cells to store one of N different values, where Nis a number of different levels that can be stored in any of the analog neural non-volatile memory cells; measuring a current drawn by the plurality of analog neural non-volatile memory cells; comparing the measured current to a target value; and identifying the plurality of the analog neural non-volatile memory cells as bad if the difference between the measured value and the target value exceeds a threshold.

Claims (20)

1 . A method comprising:

programming a first group of analog neural non-volatile memory cells in a plurality of analog neural non-volatile memory cells in an array of analog neural non-volatile memory cells with a level corresponding to the smallest cell current among N levels, where N is a number of different levels that can be stored in any of the analog neural non-volatile memory cells and where the plurality of analog neural non-volatile memory cells is fewer than all of the analog neural non-volatile memory cells in the array;

programming a second group of analog neural non-volatile memory cells in the plurality of analog neural non-volatile memory cells with a level corresponding to the largest cell current among the N levels, wherein each of the analog neural non-volatile memory cells in the second group is adjacent within the array to one or more of the analog neural non-volatile memory cells in the first group, thereby generating the highest electrical field possible among all combinations of the N levels in adjacent cells;

measuring a current drawn by the first group and the second group of analog neural non-volatile memory cells;

comparing the measured current to a target value; and

identifying the plurality of analog neural non-volatile memory cells as bad if the difference between the measured value and the target value exceeds a threshold.

2 . The method of claim 1 , wherein the plurality of analog neural non-volatile memory cells are stacked-gate flash memory cells.

3 . The method of claim 1 , wherein the plurality of analog neural non-volatile memory cells are split-gate flash memory cells.

4 . The method of claim 1 , wherein the plurality of analog neural non-volatile memory cells are part of a neural network.

5 . A system comprising:

a plurality of analog neural non-volatile memory cells in an array of analog neural non-volatile memory cells; and

test control logic to perform a checkerboard verification test comprising:

programming a first group of analog neural non-volatile memory cells in a plurality of analog neural non-volatile memory cells with a level corresponding to the smallest cell current among N levels, where N is a number of different levels that can be stored in any of the analog neural non-volatile memory cells and where the plurality of analog neural non-volatile memory cells is fewer than all of the analog neural non-volatile memory cells in the array;

programming a second group of analog neural non-volatile memory cells in the plurality of analog neural non-volatile memory cells with a level corresponding to the largest cell current among the N levels, wherein each of the analog neural non-volatile memory cells in the second group is adjacent within the array to one or more of the analog neural non-volatile memory cells in the first group, thereby generating the highest electrical field possible among all combinations of the N levels in adjacent cells;

measuring a current drawn by the first group and the second group of analog neural non-volatile memory cells;

comparing the measured current to a target value; and

identifying the plurality of analog neural non-volatile memory cells as bad if the difference between the measured value and the target value exceeds a threshold.

6 . The system of claim 5 , wherein the plurality of analog neural non-volatile memory cells are stacked-gate flash memory cells.

7 . The system of claim 5 , wherein the plurality of analog neural non-volatile memory cells are split-gate flash memory cells.

8 . The system of claim 5 , wherein the plurality of analog neural non-volatile memory cells are part of a neural network.