NAND Flash Memory Garbage Collection via Page Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current NAND flash memory systems face inefficiencies due to the need for sequential writing of pages within a block, leading to wasted storage space and reduced device lifetime from high erase operations, as well as the inability to modify data in place without erasing an entire block, which results in inefficient data management and premature wear-out.
Innovation Solution
A non-volatile data storage system with a memory module that configures storage locations as active, dead, or free, allowing data to be moved from active storage locations to free locations within the same memory area or another area, optimizing data management by reducing the need for sequential writing and minimizing erase operations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If data modification is performed by reading to volatile memory and writing to a free page, then data integrity is maintained, but the number of write operations increases and storage efficiency decreases
Solution Approach 1:
The memory system segments storage into multiple areas (first memory area, second memory area, free area) with different functions. Active pages reside in the first area, modified pages are temporarily stored in the second area, and free pages are managed separately. This segmentation allows efficient tracking of page states and reduces unnecessary operations across the entire block.
Solution Approach 2:
A page state management mechanism acts as an intermediary between the volatile memory and non-volatile memory areas. It tracks which pages are active, dead, or free, and coordinates data movement between memory areas. This intermediary layer optimizes the write process by maintaining accurate page state information without requiring full block reads or erasures.
2Reliability
If sequential writing order is enforced to minimize write-disturb errors, then reliability is improved, but storage space utilization deteriorates due to wasted free pages
Solution Approach 1:
The system dynamically manages page writing order within a block based on real-time conditions. Instead of enforcing strict sequential order, the memory controller can write to any free page within a block as long as the block's write pointer hasn't reached it, adapting the writing sequence to available space and performance needs while still preventing write-disturb errors.
Solution Approach 2:
The patent changes the parameter of page address selection from fixed sequential to flexible non-sequential within constraints. The write pointer tracks the highest address written, but new data can be placed in any free page with address less than or equal to the write pointer, optimizing space utilization while maintaining reliability through the pointer mechanism.
3Quantity of substance
If entire blocks are erased to reclaim space from dead pages, then storage is reclaimed, but device lifetime is reduced due to increased erase operations
Solution Approach 1:
The memory system segments the block into multiple areas (first memory area for active pages, second memory area for dead pages, and free area). This segmentation allows selective erasure of only the second memory area containing dead pages, rather than erasing the entire block. This reduces erase operations and extends device lifetime while effectively reclaiming storage space.
Solution Approach 2:
The patent extracts and isolates dead pages into a separate second memory area within the block. This extraction allows the system to erase only the necessary portion (second area with dead pages) to reclaim storage, rather than erasing the entire block. The first memory area with active pages remains intact, reducing unnecessary erase operations and extending device lifetime.
4Ease of operation
If in-place modification is attempted by erasing the page first, then data can be modified, but the erase operation size requirement increases significantly
Solution Approach 1:
The system uses an intermediary second memory area within the same block to store modified pages. Instead of erasing the entire block or large portions, the controller writes modified data to the second area, which is then selectively erased only when needed. This intermediary approach enables efficient in-place modification with minimal erase overhead.
Solution Approach 2:
The patent applies local quality by creating a dedicated second memory area within the block specifically for handling modifications. This localized area has different functional properties than the first memory area, allowing modifications to be confined to a small, manageable region that can be erased independently without affecting active data in the first area.
Data Source
AI summary
Non-volatile memory systems such as those using NAND FLASH technology have a property that a memory location can be written to only once prior to being erased, and a contiguous group of memory locations need to be erased simultaneously. The process of recovering space that is no longer being used for storage of current data, called garbage collection, may interfere with the rapid access to data in other memory locations of the memory system during the erase period. The effects of garbage collection on system performance may be mitigated by performing portions of the process contemporaneously with the user initiated reading and writing operations. The memory circuits and the data may also be configured such that the data is stored in stripes of a RAID array and the scheduling of the erase operations may be arranged so that the erase operations for garbage collection are hidden from the user operations.