Adaptive Read Thresholding for 3D NAND Read Disturb
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for non-volatile memory devices, such as NAND flash, face significant storage overhead due to the need for counters to track read operations, which is unacceptable for efficient data management and garbage collection, especially in scenarios where read disturb noise occurs.
Innovation Solution
Implementing a system with a memory controller that uses adaptive read thresholds and counters to determine when to perform read reclaim operations, minimizing storage overhead by selecting thresholds based on bit errors and reducing the frequency of counter updates, thereby optimizing garbage collection in 3D NAND storage systems.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If counters are used to track read operations for every page, then read disturb detection accuracy is improved, but storage overhead increases significantly
Solution Approach 1:
The patent divides the storage space into superblocks containing multiple blocks, and further divides blocks into pages. Instead of tracking read counts for every individual page, the system uses a hierarchical approach where read counts are tracked at the block level within superblocks. This segmentation allows the system to monitor read patterns effectively while reducing the total number of counters needed, thereby lowering storage overhead while maintaining detection accuracy.
Solution Approach 2:
The patent implements a universal read count tracking mechanism that serves multiple purposes: detecting read disturb, triggering garbage collection, and managing block wear. By using a single counter structure that can track reads across different pages and blocks, the system eliminates the need for separate counters for each page, reducing overall storage overhead while maintaining comprehensive monitoring capability.
2Productivity
If read counts are tracked for every page, then garbage collection efficiency is improved, but device complexity increases
Solution Approach 1:
The patent organizes storage into a hierarchical structure with superblocks containing multiple blocks, and blocks containing multiple pages. Read count tracking is implemented at the block level rather than page level, simplifying the counter management structure. This segmentation reduces the complexity of tracking and managing counters across the entire storage device while still enabling effective garbage collection decisions.
Solution Approach 2:
The patent merges the read count tracking function with the existing block management structure. Instead of maintaining separate counter structures for each page, the system combines read counting with block-level metadata management. This integration reduces device complexity by eliminating redundant counter structures while maintaining the ability to perform efficient garbage collection.
3Reliability
If adaptive read thresholds are used, then read disturb recovery is improved, but measurement precision requirements increase
Solution Approach 1:
The patent implements dynamic read thresholds that adapt based on observed bit error patterns and read disturb conditions. Instead of using fixed thresholds, the system adjusts thresholds adaptively to account for varying read disturb effects across different blocks and pages. This dynamic approach improves read disturb recovery by allowing the system to tolerate higher error rates in affected areas while maintaining strict thresholds in stable areas, effectively balancing reliability with measurement requirements.
Data Source
AI summary
A system includes memory cells arranged in blocks and a memory controller. The memory controller receives a read command to read a first block. The first block can be associated with a first read count and a first read threshold. The first read count is incremented when the first block is read, and when the first read count reaches the read threshold, a read reclaim test is performed. The first read count is set to zero after a power off or a read reclaim operation. When the first read count is zero, an adaptive read threshold is selected based on the number of bit errors. Further, in a read reclaim test, the number of bit errors is tested against an adaptive error threshold to determine whether a garbage collection operation is performed.