NAND Flash Controller Read Disturb Mitigation via Random Neighbor Selection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
NAND-type flash memory experiences data errors due to read disturb issues, leading to increased read latency overhead during sequential reading operations, which existing methods fail to adequately address.
Innovation Solution
A method of controlling a non-volatile memory system using a controller that performs random neighbor selection operations to distribute additive reading operations, reducing read latency overhead by selecting addresses for neighboring pages and optimizing reading operations based on threshold values and idle times.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If sequential reading operations are performed on neighboring pages to check for read disturb errors, then data error detection capability is improved, but read latency increases
Solution Approach 1:
The patent performs preliminary determination of whether read disturb check is necessary before executing the actual reading operation. By checking if the read count exceeds a threshold value in advance, the system avoids unnecessary sequential reading operations on neighboring pages, thus reducing read latency while maintaining data error detection capability when actually needed
Solution Approach 2:
The patent dynamically changes the read count threshold parameter based on memory usage patterns and read frequency. By adjusting this parameter, the system optimizes the balance between performing read disturb checks (improving reliability) and avoiding excessive checks (reducing time loss), adapting to different operational conditions
2Reliability
If read count threshold is set low to frequently perform read disturb checks, then data reliability is improved, but read latency overhead increases
Solution Approach 1:
The patent implements a feedback mechanism where the read count is continuously monitored and compared against the threshold. The system adjusts future read operations based on whether previous reads triggered disturb checks, creating a adaptive control loop that balances reliability and latency by learning from past operations
Solution Approach 2:
The patent makes the read disturb check mechanism dynamic by conditionally enabling it based on read count thresholds rather than performing checks on every read operation. This dynamic approach allows the system to maintain high reliability when needed while minimizing latency overhead during normal operations
Data Source
AI summary
According to example embodiments, a method of controller a memory system using a controller includes receiving a first read count command, determining if a read count of a non-volatile memory in the memory system exceeds a threshold value, and performing a first reading operation on the non-volatile memory according to the first read command. If the read count of the non-volatile memory exceeds the threshold value, then addresses are selected to which a plurality of additive reading operation corresponding to the first read command will be performed, in a random neighbor selection operation. The plurality of additive reading operations for checking data of neighboring pages of the page for performing the reading operation are distributed and processed.