NAND Memory Dispersion-Based Block Selection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing memory sub-systems face performance degradation due to inefficient garbage collection operations, which are exacerbated by not accounting for data dispersion and the type of media management operation, leading to increased power consumption and reduced lifespan in memory devices, especially in mobile deployments where space is limited.
Innovation Solution
The implementation of a data dispersion-based memory management component that determines the dispersion of valid data portions within memory blocks and selects blocks for media management operations based on these dispersions and the type of operation, improving performance by optimizing garbage collection and reducing write amplification.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If traditional garbage collection operations are performed without considering data dispersion, then memory management can be implemented, but performance degradation occurs and power consumption increases
Solution Approach 1:
The patent performs preliminary determination of data dispersion characteristics before executing garbage collection operations. By calculating the dispersion of valid data portions across memory blocks in advance, the system identifies optimal candidate blocks for media management operations, thereby avoiding unnecessary read/write operations and reducing power consumption during garbage collection.
Solution Approach 2:
The patent introduces data dispersion as a new selection parameter for garbage collection operations. Instead of using traditional methods that ignore data distribution patterns, the system changes the selection criterion to include dispersion metrics, which enables more efficient identification of candidate blocks and reduces the overall number of operations required, thus improving performance and reducing energy loss.
2Productivity
If traditional garbage collection operations are performed without considering data dispersion, then memory management can be implemented, but computational overhead and bus traffic increase
Solution Approach 1:
The patent extracts and focuses computation only on the specific metric needed for block selection—data dispersion of valid data portions. By isolating this particular calculation from other memory management computations, the system reduces overall computational overhead while still achieving improved garbage collection efficiency through informed block selection.
3Duration of action of stationary object
If memory blocks are selected without considering data dispersion, then garbage collection can be performed, but write amplification increases and device lifespan decreases
Solution Approach 1:
The patent performs preliminary analysis of data dispersion characteristics to identify optimal candidate blocks before executing write operations. By determining which blocks have valid data portions dispersed in a manner that minimizes additional writes, the system reduces write amplification and extends memory device lifespan through more efficient garbage collection operations.
Data Source
AI summary
A method includes determining a respective number of and respective locations of valid data portions of a plurality of blocks of NAND memory cells, based on the respective locations of the valid data portions, determining respective dispersions of the valid data portions within the plurality of blocks of NAND memory cells, based at least on the respective dispersions, selecting a block of NAND memory cells from the plurality of blocks of NAND memory cells, and performing a folding operation on the selected block.