Dual-Read Data Integrity Scan for Memory Sub-Systems
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current memory sub-systems face inefficiencies in data integrity checks due to inability to distinguish contributions of reliability statistics from multiple program levels, leading to unnecessary data relocation and re-programming operations, which consume system resources and cause wear on memory devices.
Innovation Solution
Implementing a dual-read data integrity scan that performs a first scan on memory cells at two program levels, followed by a second scan with a read level offset applied to one level, allowing accurate detection of reliability statistic contributions and enabling corrective actions only when necessary.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a data integrity scan is performed on memory cells at multiple program levels using a single read level, then the scan can detect overall reliability issues, but it cannot distinguish which program level contributes to the reliability statistic, leading to unnecessary data relocation and re-programming operations
Solution Approach 1:
The patent segments the data integrity scan process into multiple separate scans, each targeting a specific program level. Instead of performing a single scan that aggregates reliability statistics from multiple program levels, the system performs individual scans for each program level (e.g., first program level, second program level, third program level). This segmentation allows the system to identify which specific program level contributes to reliability issues, thereby avoiding unnecessary data relocation and re-programming operations for other program levels that are functioning correctly.
2Reliability
If the memory sub-system performs frequent data integrity scans to ensure data reliability, then detection accuracy improves, but system resources are consumed and memory device wear increases
Solution Approach 1:
The patent applies segmentation by dividing the comprehensive data integrity scan into targeted scans for specific program levels. Instead of frequently scanning all program levels regardless of their individual status, the system performs scans only when needed for specific program levels based on their individual reliability characteristics. This reduces unnecessary scanning operations, conserving system resources and reducing memory device wear while maintaining accurate detection of actual integrity issues.
Data Source
AI summary
A processing device in a memory sub-system performs a first data integrity scan on a block of a memory device to determine a first combined reliability statistic of memory cells in the block associated with a first program level and a second program level, and performs, using a predetermined read level offset corresponding to one of the first program level or the second program level, a second data integrity scan on the block of the memory device to determine a second combined reliability statistic of the memory cells in the block associated with the first program level and the second program level. The processing device determines a difference between the first combined reliability statistic and the second combined reliability statistic and, responsive to the difference between the first combined reliability statistic and the second combined reliability statistic satisfying a threshold criterion, performs a corrective action on the block of the memory device.