Channel-Scope Proximity Disturb Remapping for Non-Volatile Memory
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Non-volatile memory devices face challenges with reduced physical isolation of storage elements, leading to defects such as high read and write error rates due to thermal effects and disturb coupling, which existing solutions address at the cost of increased complexity and performance limitations.
Innovation Solution
A channel-scope proximity disturb and defect remapping scheme that remaps defective storage elements to spare locations while minimizing the introduction of new disturb relationships, using a subslice element status map and substitution data structure to manage error rates and maintain disturb relationships.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Volume of moving object
If storage elements are forced closer together to achieve smaller products, then product size is reduced, but physical isolation between storage elements decreases leading to increased disturb coupling and error rates
Solution Approach 1:
The patent divides the memory space into managed units with specific neighbor relationships, segmenting the memory into regions that can be independently remapped. This allows defective elements to be isolated and remapped to spare locations without affecting the entire memory array, thereby maintaining reliability while preserving compact physical layout.
Solution Approach 2:
The patent introduces a remapping mechanism that acts as an intermediary between logical addresses and physical storage locations. This intermediary layer (remapping logic and neighbor identification structures) decouples the physical proximity issues from logical data access, allowing defective physically-close elements to be remapped to distant spare locations while maintaining efficient access patterns.
2Reliability
If defect remapping schemes implement re-vectoring to direct logical addresses to different physical regions, then defective elements are remediated, but the complexity of identifying neighbors and the cost of remediation increases
Solution Approach 1:
The patent pre-identifies and stores neighbor relationships in data structures during manufacturing or initialization, before defects occur. This preliminary action creates a lookup table of valid neighbor relationships that can be quickly queried during operation, avoiding the need for complex real-time neighbor identification and reducing operational complexity.
Solution Approach 2:
The patent creates and maintains copies of neighbor relationship information in data structures that can be efficiently searched and updated. Instead of computing neighbor relationships on-the-fly, the system uses pre-computed copies stored in memory, reducing the computational complexity of neighbor identification during defect remapping operations.
3Reliability
If remapping expands the set of neighbor units, then more defective elements can be remediated, but the cost of remediation and complexity of managing expanded neighbor sets increases
Solution Approach 1:
The patent implements dynamic remapping where the set of active neighbors and their relationships can be adjusted based on defect patterns. The system can dynamically expand or contract the effective neighbor set depending on which elements are defective, optimizing the balance between defect coverage and management complexity rather than using a fixed large neighbor set.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
Effectively remaps defective storage elements to spare locations, maintaining disturb relationships and improving data reliability with reduced complexity and cost, enabling efficient error management across multiple dice.
Implementation Method 1
For media that use a thermal process to encode the state of the storage element, such as phase change media, thermal effects propagate to surrounding storage elements. This document refers to these propagated thermal effects as disturb coupling.
Data Source
AI summary
Techniques for remapping portions of a plurality of non-volatile memory (NVM) dice forming a memory domain. A processing device partitions each NVM die into subslice elements comprising respective physical portions of NVM having proximal disturb relationships. The NVM allocation has user subslice elements and spare subslice elements. For the NVM dice forming the memory domain, the processing device performs an error analysis to identify a predetermined number of subslice elements having highest error rates for the memory domain. Identified user subslice elements having the highest error rates, remap to spare subslice elements of the memory domain that were not identified as having the highest error rates to remove subslice element or elements having highest error rates. At least one user subslice element is remapped from a first die of the memory domain to a second die of the memory domain.