ECC-Guided Storage Equalizer Tuning for Inter-Cell Interference

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing storage devices face challenges in addressing inter-cell interference (ICI) due to voltage leakage between memory cells, which affects data integrity and requires equalization methods that are not adaptable to changes in device characteristics over time.

Innovation Solution

A tunable and adaptable equalization process using machine-learning assisted voltage equalization that updates threshold models based on valid data from different sectors of the storage device, enhancing data accuracy and reliability.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If equalization is based on characteristics at a single workpoint, then the equalization process is simple and fast, but it cannot adapt to changes in device characteristics over time

Engineering Contradiction:
Improveadaptability to device characteristic changesVSAvoidequalization process complexity
Core Design Contradiction:
Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The equalization process transitions from static (single workpoint) to dynamic (multiple workpoints) by implementing a feedback mechanism that continuously monitors read validity and updates threshold models accordingly. The system adapts its equalization parameters based on actual device performance at different workpoints, making the equalization process dynamic rather than fixed.

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The system implements feedback by monitoring whether read data is valid or invalid, and using this information to determine whether to update the threshold models. When invalid data is detected, the system retrieves reference data, compares it with expected values, and updates the equalization parameters accordingly, creating a closed-loop feedback system that continuously optimizes performance.

Inventive Principle:
Principle #23Feedback

2Measurement precision

If equalization is performed frequently to maintain accuracy, then data reading accuracy improves, but processing time increases

Engineering Contradiction:
Improvedata reading accuracyVSAvoidprocessing time
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

Instead of performing full equalization updates continuously, the system applies partial updates only when necessary - specifically when invalid data is detected during normal read operations. This selective updating approach maintains accuracy when needed while avoiding unnecessary processing overhead during normal operation.

Inventive Principle:
Principle #16Partial or excessive action

Solution Approach 2:

The system changes its operation mode based on detected conditions: during normal operation it uses existing threshold models for fast reading, but when invalid data is detected, it transitions to updating parameters by retrieving reference data and recalibrating threshold models, thus adapting the processing intensity to actual needs.

Inventive Principle:
Principle #35Parameter changes

Data Source

PatentEP4664462A1Equalizer with tunable configuration according to ECC output
Publication Date: 2025.12.17 SAMSUNG ELECTRONICS CO LTD
  • EP4664462A1 patent drawingFigure 1
  • EP4664462A1 patent drawingFigure 2
  • EP4664462A1 patent drawingFigure 3

AI summary

Methods and devices for controlling a storage device including a non-volatile memory including a plurality of sectors including a first sector and a second sector; and a storage controller configured to: read first data from the first sector based on a plurality of threshold models; based on determining that the first data is invalid, read second data corresponding to the second sector based on the plurality of threshold models; based on determining that the second data is valid, update at least one parameter of the plurality of threshold models based on the second data; and generate updated first data corresponding to the first sector based on the plurality of threshold models.