Memory Controller Multi-Cluster Read Thresholds
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Memory systems face challenges in efficiently performing read operations due to shifted read threshold voltages, leading to errors and varying optimal read thresholds across different pages, which existing technologies struggle to address effectively.
Innovation Solution
A memory system and method that utilize multiple clusters of read threshold values to split pages into groups and perform additional read operations, generating new clusters based on fail bit differences to adapt and optimize read retry operations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If multiple read operations are performed using a single cluster of read threshold voltages, then the read operation can be completed with a simple threshold structure, but the read accuracy deteriorates when threshold voltages shift across different pages
Solution Approach 1:
The patent divides the memory pages into multiple groups, with each group assigned to a specific cluster of read threshold voltages. This segmentation allows different threshold clusters to be optimized for different page groups, improving read accuracy for each group while managing overall system complexity through organized division.
Solution Approach 2:
The patent implements dynamic selection of read threshold clusters based on the detected fail bit patterns. When a first cluster fails to achieve sufficient read accuracy (fail bit difference exceeds threshold), the system dynamically switches to a second cluster. This dynamic adaptation enables the system to respond to varying threshold voltage conditions across different pages, improving overall read reliability.
2Productivity
If a single cluster of read threshold voltages is used for all pages, then the device complexity is reduced, but the read operation time increases due to retries and error correction
Solution Approach 1:
The patent performs preliminary classification of pages into groups and assigns appropriate threshold clusters before executing read operations. By pre-organizing pages into groups that correspond to specific threshold clusters, the system avoids the need for multiple retry attempts with a single cluster, thereby improving read operation speed while managing complexity through advance preparation.
Solution Approach 2:
The patent incorporates feedback mechanisms that monitor fail bit patterns during read operations. When the difference between maximum and minimum fail bits exceeds a threshold, the system uses this feedback to switch from the first cluster to the second cluster. This feedback-driven approach enables the system to adapt to varying conditions and improve read speed by avoiding unnecessary retries.
3Reliability
If read threshold voltages are fixed for all pages, then the operation is simple and fast, but the data access reliability deteriorates when optimal thresholds vary across pages
Solution Approach 1:
The patent applies the principle of local quality by assigning different read threshold voltage characteristics to different page groups based on their specific requirements. Each cluster is optimized for the local characteristics of its assigned page group, allowing the system to achieve high data access reliability for each group while minimizing the overall time penalty through targeted rather than universal optimization.
Data Source
AI summary
A memory system includes a memory device and a controller. The controller performs multiple read operations on a target block, using a first duster of read threshold voltages. The controller generates a second duster of read threshold voltages using the first cluster when a difference between the maximum number of fail bits and the minimum number of fail bits associated with the multiple read operations exceeds a threshold. The controller splits pages in the target block into a first group of pages for the first cluster and a second group of pages for the second cluster. The controller performs additional read operations on the first group of pages using the first cluster and on the second group of pages using the second cluster.