Half-Good Memory Block Management With Deck-Level Tracking
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing memory sub-systems inefficiently manage half-good blocks, leading to reduced overprovisioning, decreased system performance, and increased wear on remaining valid blocks due to the binary classification of blocks as fully good or fully bad, which fails to utilize partially defective blocks effectively.
Innovation Solution
Implementing bad-deck management by using a pointer to track half-blocks and generating a bitmap to determine the health status of individual decks within a block, allowing for the independent management and utilization of partially good decks, thereby increasing the number of valid blocks available for storage.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If blocks are classified binary as fully good or fully bad, then management is simple, but storage efficiency decreases and overprovisioning is reduced
Solution Approach 1:
The patent segments a block into multiple decks (e.g., upper deck and lower deck), allowing independent health status evaluation of each deck. This enables partial utilization of blocks where only some decks are defective, thereby increasing the number of usable storage units without complicating the overall management structure.
Solution Approach 2:
The patent applies local quality by assigning different health statuses to different decks within the same block. Instead of treating the entire block uniformly, each deck can be independently marked as good or bad, allowing the system to utilize healthy portions of blocks while isolating defective regions.
2Reliability
If half-good blocks are discarded entirely, then reliability is maintained, but waste increases and overprovisioning decreases
Solution Approach 1:
By segmenting blocks into decks, the patent allows the system to discard only the defective decks while retaining and utilizing the healthy decks. This segmentation prevents the complete discarding of half-good blocks, thereby reducing storage capacity waste while maintaining data reliability through proper isolation of bad decks.
Solution Approach 2:
The patent converts the presence of defective decks into a benefit by using them to identify and isolate only the specific bad portions, thereby transforming potential harm (defective blocks) into useful information for optimizing storage utilization and reducing waste.
3Productivity
If deck-level tracking is implemented, then storage efficiency increases, but system complexity increases
Solution Approach 1:
The patent introduces a deck-level tracking mechanism that segments block management into finer granular units. While this increases storage efficiency by enabling partial utilization, it also increases system complexity by requiring additional data structures (e.g., deck-level bitmaps) and management logic to track the health status of individual decks within blocks.
Data Source
AI summary
A system includes a memory device and a processing device operatively coupled to the memory device. The processing device is to perform operations including setting a partial translation unit (TU) pointer to identify a first partial-TU of an ordered sequence of partial-TUs, the ordered sequence spanning over a plurality of dies of the memory device. The operations further include, responsive to determining that the first partial-TU has a good health status, appending the first partial-TU to a partial-TU stripe and incrementing the partial-TU pointer to identify a second partial-TU. The operations further include, responsive to determining that the second partial-TU has a bad health status, incrementing the partial-TU pointer without adding the second partial-TU to the partial-TU stripe. The operations further include performing one or more write operations on a plurality of TUs comprised by the partial-TU stripe.