IP Library › Granted Patent US 12,738,322
Granted Patent B2
US 12,738,322 · App. 18/897,326 · Granted Sep 15, 2026

Apparatus and methods for modifying impedances in time-based vector-matrix multiplication circuits

Inventors: Mohammad Mahmoodi (Santa Barbara, CA); Zahra Fahimi (Santa Barbara, CA); Martin Lueker-Boden (Fremont, CA)
Assignee: Sandisk Technologies, Inc.
G11C16/102G11C5/063G11C16/24
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 12,738,322
App. No.
18/897,326
Filed
Sep 26, 2024
Granted
Sep 15, 2026
Kind
B2
Art Unit
2827
USPC
365/189.011
Abstract

An apparatus is provided that includes a memory array including non-volatile memory cells configured to store weights of an artificial neural network, a bit line coupled to a plurality of the non-volatile memory cells, and an interface circuit coupled to the bit line, the interface circuit configured to amplify an output impedance of the memory array.

Claims (36)

1 . An apparatus comprising:

a memory array comprising non-volatile memory cells configured to store weights of an artificial neural network;

a bit line coupled to a plurality of the non-volatile memory cells; and

an interface circuit coupled to the bit line, the interface circuit configured to amplify an output impedance of the memory array.

2 . The apparatus of claim 1 , wherein an output impedance of the interface circuit is greater than an output impedance of the memory array.

3 . The apparatus of claim 1 , wherein an output impedance of the interface circuit substantially equals a multiple of the output impedance of the memory array.

4 . The apparatus of claim 1 , wherein the interface circuit comprises a transistor comprising a cascode configuration.

5 . The apparatus of claim 1 , wherein the interface circuit comprises an amplifier.

6 . The apparatus of claim 5 , wherein the interface circuit comprises:

an amplifier comprising an input terminal coupled to the bit line; and

a transistor comprising a source terminal coupled to the bit line, and a gate terminal coupled to an output terminal of the amplifier.

7 . The apparatus of claim 1 , further comprising a sense amplifier coupled to the interface circuit.

8 . The apparatus of claim 7 , wherein the sense amplifier comprises a capacitor configured to integrate a current from the bit line.

9 . The apparatus of claim 7 , wherein the sense amplifier comprises a resistive circuit element configured to reduce an input impedance of the sense amplifier.

10 . The apparatus of claim 7 , wherein the sense amplifier comprises a capacitor and a resistive circuit element disposed in parallel with the capacitor.

11 . The apparatus of claim 7 , wherein the sense amplifier comprises a capacitor and a transistor disposed in parallel with the capacitor, the transistor biased to operate in an ohmic region.

12 . The apparatus of claim 1 , wherein the memory array is configured to perform vector matrix multiplication.

13 . The apparatus of claim 1 , wherein the memory array comprises a plurality of 3D NAND memory strings coupled to the bit line.

14 . The apparatus of claim 1 , comprising a time-based vector-matrix multiplication circuit.

15 . An apparatus comprising

an array of vertical NAND strings, each comprising a plurality of non-volatile memory cells, each non-volatile memory cell configured to conduct a current based on a corresponding weight of a neuromorphic computing system;

a plurality of bit lines, each coupled to a corresponding plurality of the vertical NAND strings; and

a plurality of sense amplifiers, each coupled to a corresponding bit line,

wherein each sense amplifier comprises:

a capacitor configured to integrate a current conducted by the corresponding bit line; and

a resistive circuit element configured to reduce an input impedance of the sense amplifier.

16 . The apparatus of claim 15 , wherein the resistive circuit element comprises a resistor disposed in parallel with the capacitor.

17 . The apparatus of claim 15 , wherein the resistive circuit element comprises a transistor disposed in parallel with the capacitor.

18 . The apparatus of claim 15 , wherein each sense amplifier comprises an input impedance substantially equal to a resistance of the resistive circuit element.

19 . The apparatus of claim 15 , comprising a time-based vector-matrix multiplication circuit.

20 . A method comprising:

performing a time-based vector-matrix multiplication by:

programming a current in each memory cell of a 3D NAND memory array, the 3D NAND memory array comprising a first output impedance and a plurality of NAND strings coupled to a bit line;

coupling a cascode-configured transistor to the bit line, the cascode-configured transistor comprising a second output impedance greater than the first output impedance; and

coupling the cascode-configured transistor to an input of a sense amplifier comprising an integration capacitor disposed in parallel with a resistor,

wherein an input resistance of the sense amplifier substantially equals a resistance of the resistor.

Assignments (4)
PARTIAL RELEASE OF SECURITY INTERESTS Recorded Apr 25, 2025
From: JPMORGAN CHASE BANK, N.A., AS AGENT
To: SANDISK TECHNOLOGIES, INC.
Reel/Frame 071382/0001 →
SECURITY AGREEMENT Recorded Apr 25, 2025
From: SANDISK TECHNOLOGIES, INC.
To: JPMORGAN CHASE BANK, N.A., AS COLLATERAL AGENT
Reel/Frame 071050/0001 →
SECURITY AGREEMENT (SUPPLEMENTAL) Recorded Nov 14, 2024
From: SANDISK TECHNOLOGIES, INC.
To: JPMORGAN CHASE BANK, N.A., AS COLLATERAL AGENT
Reel/Frame 069411/0486 →
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Sep 26, 2024
From: MAHMOODI, MOHAMMAD; FAHIMI, ZAHRA; LUEKER-BODEN, MARTIN
To: SANDISK TECHNOLOGIES, INC.
Reel/Frame 068707/0189 →
Continuity (1)
Related Publication 20260088093A1 · Mar 26, 2026
References Cited (13)
US 6307791B1 · Otsuka · 2001 [cited by examiner]
US 7126853B2 · Kim · 2006 [cited by applicant]
US 7751251B2 · Bauer · 2010 [cited by applicant]
US 10411704B1 · Lin · 2019 [cited by examiner]
US 10635398B2 · Lin et al. · 2020 [cited by applicant]
US 11354562B2 · Tran et al. · 2022 [cited by applicant]
US 20050002223A1 · Coteus · 2005 [cited by examiner]
US 20210019609A1 · Strukov et al. · 2021 [cited by applicant]
US 20230282288A1 · Yang · 2023 [cited by examiner]
US 20240313776A1 · Watanabe · 2024 [cited by examiner]
US 20240379136A1 · Yeh · 2024 [cited by examiner]
CN 114464240A · 2022 [cited by examiner]
Mohammad Bavandpour et al., “3D-aCortex: an ultra-compact energy-efficient neurocomputing platform based on commercial 3D-NAND flash memories”, Neuromorphic Computer Engineering, Jul. 15, 2021. [cited by applicant]