Open-Block Wordline Outlier Band Detection for Data Retention
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing data storage devices face challenges in detecting and correcting read voltage shifts in open blocks of memory due to temperature changes and aging, leading to increased bit error rates and reduced data retention, especially in multi-meta die environments with large block sizes, which can result in significant data loss and performance degradation.
Innovation Solution
The implementation of a data storage device that groups continuous wordlines crossing a data retention threshold as a band and classifies them as an outlier band using a second derivative approach for early detection of bit error rate regions, allowing for proactive correction of affected blocks.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If continuous monitoring of all wordlines in open blocks is performed to detect read voltage shifts, then data retention and reliability are improved, but device complexity and computational overhead increase significantly
Solution Approach 1:
The patent segments the wordline space into multiple bands and uses sampling points within each band to detect read voltage shifts. Instead of monitoring all wordlines individually, the system divides the monitoring task into manageable segments (bands) and selects representative sampling points, significantly reducing computational complexity while maintaining detection effectiveness.
Solution Approach 2:
The system performs preliminary classification of bands into outlier bands and non-outlier bands based on sampling points before full monitoring is needed. This preliminary action identifies which bands require detailed monitoring and which can be monitored less frequently, reducing overall system complexity while maintaining reliability.
2Measurement precision
If a threshold-based classification method is used to identify outlier bands, then detection precision is improved, but the difficulty of detecting and measuring increases due to the need for second derivative calculations
Solution Approach 1:
The patent introduces sampling points as intermediaries between the raw wordline data and the outlier band classification. These sampling points serve as mediators that capture the essential characteristics of each band without requiring direct analysis of all wordlines, simplifying the measurement process while maintaining precision through the threshold-based classification of bands.
3Reliability
If proactive correction of outlier bands is implemented, then bit error rate is reduced and read performance is improved, but loss of time increases due to additional processing steps
Solution Approach 1:
The system performs preliminary identification and classification of outlier bands before they cause significant bit errors. By detecting and flagging problematic bands early through sampling point analysis, the system can apply corrective measures proactively, reducing the overall processing time needed for error correction compared to reactive approaches that wait for errors to manifest.
Solution Approach 2:
The patent applies correction measures locally to identified outlier bands rather than performing global correction across all data. This localized approach focuses computational resources only on the specific bands that require correction, minimizing the time loss associated with processing while effectively reducing bit error rates in the affected regions.
Data Source
AI summary
A data storage device groups a set of continuous wordlines crossing a data retention threshold together as a band. The set of continuous wordlines are in an open block in the memory. The data storage device then determines whether a number of continuous wordlines in the set is greater than a threshold. If the number of continuous wordlines in the set is greater than the threshold, the data storage device classifies the set of continuous wordlines as an outlier band in the open block.