Dynamic Block Mode Conversion in Non-Volatile Memory
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The static assignment of blocks to different tiers in multi-tiered memory storage systems can lead to premature capacity exhaustion or endurance limits, as blocks are not dynamically adjusted based on bit error rates and wear metrics.
Innovation Solution
A non-volatile memory system that dynamically converts physical blocks between QLC and SLC modes based on bit error rates, write stream metrics, and block health, allowing for the transfer of blocks with higher bit error rates to SLC mode and healthier blocks to QLC mode, optimizing storage capacity and endurance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If blocks are statically assigned to QLC or SLC tiers, then storage capacity is optimized, but block endurance is reduced due to premature capacity exhaustion or wear limits
Solution Approach 1:
The patent implements dynamic block mode conversion where blocks can transition between QLC and SLC modes based on real-time wear metrics and bit error rates. The flash controller continuously monitors block health and automatically converts blocks from QLC to SLC mode when wear thresholds are exceeded, and vice versa when blocks are refreshed. This dynamic adjustment resolves the contradiction by allowing the system to maintain optimal storage capacity while extending block endurance through adaptive mode switching.
Solution Approach 2:
The patent changes the operational parameter of blocks from fixed mode to convertible mode between QLC and SLC. By monitoring bit error rates and wear metrics, the system dynamically adjusts the storage mode parameter of individual blocks, enabling them to operate in the mode most suitable for their current health state. This parameter change allows the same physical block to provide different storage capacities at different lifecycle stages, resolving the contradiction between capacity optimization and endurance extension.
2Quantity of substance
If blocks operate in QLC mode, then storage density increases, but bit error rate increases due to higher wear
Solution Approach 1:
The patent applies local quality by allowing different blocks to operate in different modes (QLC or SLC) based on their individual wear characteristics and bit error rates. Instead of uniformly operating all blocks in QLC mode for maximum density, the system identifies specific blocks that have accumulated excessive wear or exhibit high bit error rates and converts only those blocks to SLC mode. This localized mode conversion maintains high storage density for healthy blocks while ensuring reliable operation for worn blocks, resolving the contradiction between storage density and bit error rate.
Solution Approach 2:
The patent implements feedback mechanisms where the flash controller continuously monitors bit error rates and wear metrics of blocks operating in QLC mode. When feedback indicates that a block's bit error rate exceeds acceptable thresholds, the system automatically triggers mode conversion from QLC to SLC for that specific block. This closed-loop feedback system ensures that storage density is maximized while maintaining acceptable bit error rates by dynamically adjusting block modes based on real-time performance data.
3Duration of action of stationary object
If blocks are converted between modes dynamically, then device lifetime is extended, but system complexity increases
Solution Approach 1:
The patent implements self-service by enabling the flash controller to automatically monitor block wear metrics, determine when mode conversion is necessary, and execute the conversion process without external intervention. The system autonomously manages the complexity of dynamic mode conversion by incorporating wear tracking, threshold comparison, and mode switching logic into the flash controller's existing firmware. This self-managing approach extends device lifetime through adaptive block mode conversion while minimizing the operational complexity burden on users or external systems.
Data Source
AI summary
A non-volatile memory includes a plurality of physical blocks of storage each including a respective plurality of cells, where each of the plurality of cells is individually capable of storing multiple bits of data. A controller assigns physical blocks among the plurality of physical blocks to a first pool containing physical blocks operating in a first (e.g., QLC) mode for storing a greater number of bits per cell and assigns other physical blocks among the plurality of physical blocks to a second pool containing physical blocks operating in a second (e.g., SLC) mode for storing a lesser number of bits per cell. The controller transfers physical blocks between the first pool and the second pool based on at least bit error rates measured for the transferred physical blocks.