Non-Volatile Memory Vertical Re-Vectoring for Defect Uniformity
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Non-volatile memory systems suffer performance degradation due to uneven distribution of defective blocks across bands, leading to inconsistent system performance as some bands have significantly more defective blocks than others.
Innovation Solution
The system implements a vertical re-vectoring process to redistribute defective blocks from one band to non-defective blocks within the same plane or die's plane, ensuring that all bands have a consistent number of defects, thereby improving overall system performance by allocating defective blocks to a dead region.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If defective blocks are distributed unevenly across bands, then manufacturing precision is reduced, but device complexity remains low
Solution Approach 1:
The memory device is divided into multiple bands, each containing multiple blocks. This segmentation allows defective blocks to be isolated and managed at the band level, improving defect distribution uniformity without requiring complete redesign of the entire memory structure.
Solution Approach 2:
The patent introduces a hierarchical organization where bands are grouped into superbands and further organized across multiple planes. This multi-dimensional structure allows defective blocks to be redistributed across different bands within the same plane or across planes, achieving more uniform defect distribution while maintaining manageable complexity through structured organization.
2Productivity
If vertical re-vectoring is performed to redistribute defective blocks, then system performance is improved, but device complexity increases
Solution Approach 1:
The system performs preliminary identification and cataloging of defective blocks during manufacturing or initial operation. This preliminary action allows the control logic to pre-establish re-vectoring mappings, so that when performance optimization is needed, the system can quickly activate pre-planned redistributions rather than performing complex real-time analysis.
Solution Approach 2:
The control logic continuously monitors system performance and defect distribution across bands. Based on this feedback, the system dynamically adjusts re-vectoring operations to optimize performance. The feedback mechanism enables the system to identify when re-vectoring is needed and to evaluate its effectiveness, improving system performance through iterative optimization.
3Ease of operation
If defective blocks are concentrated in specific bands, then ease of operation is reduced, but loss of substance is minimized
Solution Approach 1:
The patent merges multiple bands into superbands and organizes them across multiple planes, creating a unified memory space where defective blocks in any single band do not isolate that entire region. This merging allows the system to access usable blocks across different bands more easily, improving ease of operation while maintaining maximum utilization of available memory capacity.
Solution Approach 2:
The system dynamically changes the logical-to-physical mapping parameters through re-vectoring operations. When defects are identified in certain bands, the control logic modifies addressing parameters to redirect accesses away from defective blocks to healthy blocks in other bands, thereby improving ease of operation without losing usable capacity.
Data Source
AI summary
Systems and methods are disclosed for improving performance of a system having non-volatile memory (“NVM”). The system can vertically re-vector defective blocks of a user region of the NVM to other blocks having the same plane or die's plane (“DIP”) but corresponding to a dead region of the NVM. Then, the system can select any band with more than one defective block and vertically re-vector one of its defective blocks to a band that has no defective blocks. At run-time, the system can monitor the number of vertical re-vectors per DIP. If at least one vertical re-vector has been performed on all DIPs of the NVM, a band of the user region can be allocated for the dead region.