Memory Controller Read Disturb Detection Using Bit-Flip Statistics

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing memory technologies face challenges in efficiently identifying and addressing read disturb states in NAND-type memory devices, leading to bit flipping and data loss due to complex decoding processes and high power consumption during read operations.

Innovation Solution

A memory controller and system that determines read disturb states by analyzing statistical results of bit flipping and unsuccessful reads at specific read voltages, using simple comparison methods with preset thresholds, reducing the need for complex decoding and minimizing power consumption.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If complex decoding processes are used to identify read disturb states, then measurement precision is improved, but device complexity and power consumption increase

Engineering Contradiction:
Improveread disturb state identification accuracyVSAvoiddecoding process complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent extracts only the essential statistical features (bit flipping count and unsuccessful read count) from the read results, eliminating the need for complex decoding processes. By taking out only the necessary information for read disturb detection, the system achieves accurate identification while significantly reducing computational complexity and power consumption.

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The patent uses simple statistical counting methods instead of expensive complex decoding algorithms. The read results are processed through lightweight statistical analysis that consumes minimal power and computational resources, effectively replacing heavy decoding operations with economical statistical evaluation.

Inventive Principle:
Principle #27Cheap short-living objects (Disposable)

2Measurement precision

If complex decoding processes are used to identify read disturb states, then measurement precision is improved, but power consumption increases

Engineering Contradiction:
Improveread disturb state identification accuracyVSAvoidpower consumption during read operations
Core Design Contradiction:
Measurement precisionVSUse of energy by moving object

Solution Approach 1:

The patent extracts only the essential statistical features (bit flipping count and unsuccessful read count) from the read results, eliminating the need for complex decoding processes. By taking out only the necessary information for read disturb detection, the system achieves accurate identification while significantly reducing computational complexity and power consumption.

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The patent uses simple statistical counting methods instead of expensive complex decoding algorithms. The read results are processed through lightweight statistical analysis that consumes minimal power and computational resources, effectively replacing heavy decoding operations with economical statistical evaluation.

Inventive Principle:
Principle #27Cheap short-living objects (Disposable)

3Productivity

If statistical analysis is performed on read results to identify read disturb states, then productivity is improved, but device complexity increases

Engineering Contradiction:
Improveread disturb detection speedVSAvoidstatistical analysis complexity
Core Design Contradiction:
ProductivityVSDevice complexity

Solution Approach 1:

The patent extracts only the essential statistical features (bit flipping count and unsuccessful read count) from the read results, eliminating the need for complex decoding processes. By taking out only the necessary information for read disturb detection, the system achieves accurate identification while significantly reducing computational complexity and power consumption.

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The patent uses simple statistical counting methods instead of expensive complex decoding algorithms. The read results are processed through lightweight statistical analysis that consumes minimal power and computational resources, effectively replacing heavy decoding operations with economical statistical evaluation.

Inventive Principle:
Principle #27Cheap short-living objects (Disposable)

Data Source

PatentUS20260037428A1Memory controller, device, system, operating method thereof, and storage medium
Publication Date: 2026.02.05 YANGTZE MEMORY TECH CO LTD
  • US20260037428A1 patent drawing
  • US20260037428A1 patent drawing
  • US20260037428A1 patent drawing

AI summary

The embodiment of the present disclosure provides a memory controller, device, system, and method. The memory controller includes an interface and a processor, wherein the processor is configured to: obtain a statistical result of a physical page corresponding to a first read voltage, wherein the statistical result includes at least one of a first number of bits flipped in two read results of the physical page corresponding to the first read voltage and a read voltage whose voltage difference with the first read voltage is less than a preset voltage, or a second number of bits of the physical page not successfully read using the first read voltage; the first read voltage is less than a mean value of a threshold voltage interval corresponding to an intermediate storage state of the physical page; and determine a read disturb state of the physical page according to the statistical result.