Memory Controller Candidate Data Selection for Write Disturbance
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Memory systems face data reliability issues due to the write disturbance effect, where data corruption occurs in one location due to writing data at another location, leading to errors.
Innovation Solution
A memory system with a non-volatile memory device and a controller that generates candidate data sets, determines the number of vulnerable cells for each set, and selectively stores the set with the smallest number of vulnerable cells to minimize write disturbance effects, thereby improving data reliability.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If data is written to a memory location, then the data is stored successfully, but write disturbance effect causes data corruption in neighboring locations
Solution Approach 1:
The controller performs preliminary actions by generating multiple candidate data sets and determining vulnerable cells before actually writing data to memory. It selects the optimal candidate that minimizes write disturbance effects on neighboring memory regions, thereby preventing data corruption before it occurs.
Solution Approach 2:
The patent converts the harmful write disturbance effect into a beneficial selection criterion. By identifying which candidate data sets produce fewer vulnerable cells (cells susceptible to write disturbance), the system uses the potential harm as a metric to select the best writing strategy, turning a negative effect into a positive selection mechanism.
2Reliability
If multiple candidate data sets are generated and evaluated, then data reliability is improved, but processing time and complexity increase
Solution Approach 1:
The controller segments the data writing process into distinct phases: generating multiple candidate data sets, evaluating each candidate's vulnerable cell count, and selecting the optimal candidate. This segmentation allows systematic evaluation of different writing strategies without overwhelming complexity, as each phase handles a specific sub-task.
Solution Approach 2:
The system changes the parameter being optimized from simple data storage to minimizing vulnerable cell count across multiple candidate data sets. By evaluating candidates based on their vulnerable cell characteristics rather than just storage efficiency, the system finds an optimal balance between reliability and processing complexity through parameter-based selection.
3Reliability
If vulnerable cells are identified and avoided, then write disturbance errors are reduced, but additional processing steps are required
Solution Approach 1:
The controller performs preliminary identification of vulnerable cells associated with each candidate data set before the actual memory write operation. By determining which cells would be vulnerable to write disturbance in advance, the system avoids time-consuming error correction or re-writing operations, reducing overall processing time despite the additional evaluation step.
Solution Approach 2:
The system skips potentially problematic candidate data sets that would create excessive vulnerable cells by selecting only the optimal candidate with minimum vulnerable cell count. This skipping approach avoids time-consuming operations on suboptimal candidates, rushing through the selection process to reach the best solution efficiently.
Data Source
AI summary
A memory system includes a non-volatile memory device and a controller. The non-volatile memory device includes a plurality of memory regions, each memory region including a plurality of cells commonly coupled to a word line. The controller generates a plurality of candidate data sets based on source data, determines a number of vulnerable cells corresponding to each of the plurality of candidate data sets, and stores a candidate data set having a smallest number of vulnerable cells into a target memory region among the plurality of memory regions.