Memory Subsystem Error Handling via Mitigation Block Remapping
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional memory sub-systems face challenges in error handling, particularly with data migration in zone-based memory devices, leading to increased write amplification and overprovisioning requirements, as well as the lack of redundant data protection, which can result in data corruption and loss during error events.
Innovation Solution
The solution involves copying data from a source block experiencing an error to a spare mitigation block within the same plane, allowing for remapping of the source block's identifier to the mitigation block, thereby redirecting read requests and ensuring data integrity without migrating the entire block set, thus reducing write amplification and overprovisioning needs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If data is migrated from a source block to a mitigation block when an error event is detected, then data integrity is protected, but write amplification increases
Solution Approach 1:
The patent segments the data migration process by copying only the specific data from the source block to the mitigation block when an error is detected, rather than migrating entire block sets. This selective segmentation reduces the volume of write operations and consequently lowers write amplification while still protecting data integrity.
Solution Approach 2:
The patent applies partial action by performing data migration only for the affected source block when an error event occurs, rather than preemptively migrating all blocks. This partial approach minimizes unnecessary write operations and reduces write amplification while maintaining sufficient data protection.
2Reliability
If redundant data protection mechanisms are implemented, then data loss prevention is improved, but device complexity increases
Solution Approach 1:
The patent introduces a mitigation block as an intermediary structure that simplifies error handling. When an error is detected in a source block, the system copies data to the mitigation block and updates the block set map, providing a straightforward remediation path without requiring complex redundant protection mechanisms across the entire system.
Solution Approach 2:
The patent applies local quality by implementing error protection specifically at the source block level where errors are detected, rather than applying uniform redundant protection across all blocks. This localized approach maintains data loss prevention while reducing overall system complexity.
3Reliability
If entire block sets are migrated for error protection, then data safety is ensured, but overprovisioning requirements increase
Solution Approach 1:
The patent segments the protection scope from entire block sets down to individual source blocks. By copying only the necessary data from affected source blocks to mitigation blocks, the system ensures data safety while significantly reducing the amount of additional storage capacity required, thereby lowering overprovisioning requirements.
Solution Approach 2:
The patent uses partial action by migrating only the specific data from source blocks that experience errors, rather than preemptively migrating entire block sets. This partial migration approach maintains data safety while minimizing the storage overhead and reducing overprovisioning requirements.
Data Source
AI summary
A system including a memory device having blocks of memory cells and a processing device operatively coupled to the memory device. The processing device to perform operations comprising: detecting an error event triggered within a source block of the memory cells; reading data from the source block; writing the data into a mitigation block that is different than the source block; and replacing, in a mapping data structure, a first identifier of the source block with a second identifier of the mitigation block.