Memory Health Monitoring via Shaped Data Alternation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing memory health monitoring systems fail to accurately assess the health of memory blocks due to the uneven wear caused by shaped data, which can mask bit failures, leading to inaccurate bit error rates and premature determination of block health metrics.
Innovation Solution
The system alternates between writing shaped data and de-shaped data to memory blocks during different P/E cycles, with de-shaped data being written during polling cycles to provide a more accurate block health metric, and shaped data being written during other cycles to reduce wear.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Duration of action of stationary object
If shaped data is continuously written to memory blocks to reduce wear, then the memory block lifetime is extended, but the bit error rate measurement becomes inaccurate and block health assessment deteriorates
Solution Approach 1:
The system implements periodic alternation between writing shaped data and de-shaped data to memory blocks. During polling cycles, de-shaped data is written to enable accurate bit error rate measurement, while during other cycles shaped data is written to reduce wear. This periodic switching resolves the contradiction by ensuring both wear reduction and measurement accuracy are maintained at different intervals.
Solution Approach 2:
The system performs preliminary action by writing de-shaped data to memory blocks before polling cycles occur. This preparatory step ensures that when health assessment is needed, the memory block contains appropriate data patterns that enable accurate bit error rate measurement, thus preventing measurement inaccuracy before it occurs.
2Object-affected harmful factors
If shaped data with more logical one values is stored to reduce wear, then wear on memory blocks is reduced, but some bit failures remain undetected leading to inaccurate health metrics
Solution Approach 1:
The system periodically switches between storing shaped data and de-shaped data. When shaped data is stored, wear is reduced. When de-shaped data is stored during polling cycles, comprehensive bit failure detection is enabled. This periodic alternation ensures both wear reduction and reliable health assessment are achieved at appropriate intervals.
Solution Approach 2:
The system changes the data pattern parameter (from shaped data with more logical one values to de-shaped data with more balanced distribution) before polling cycles. This parameter change enables the memory system to transition from a wear-reduction state to a measurement-accuracy state, ensuring both wear reduction and reliable bit failure detection are achieved.
Data Source
AI summary
A data storage device may be configured to write first data to a first set of storage elements of a non-volatile memory and to write second data to a second set of storage elements of the non-volatile memory. The first data may be processed by a data shaping operation, and the second data may not be processed by the data shaping operation. The data storage device may be further configured to read a representation of the second data from the second set of storage cells and to determine a block health metric of a portion of the non-volatile memory based on the representation of the second data. The portion may include the first set of storage elements and the second set of storage elements. As an illustrative, non-limiting example, the first portion may be a first block of the non-volatile memory.