NAND Threshold Voltage Modulation for Inter-Cell Interference
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Flash memory devices suffer from inter-cell interference (ICI) which affects the endurance and reliability of memory cells, limiting their program-erase cycles and data retention time, despite the use of advanced error-correction codes.
Innovation Solution
Utilizing artificial intelligence (AI) to optimize threshold voltage targets for memory cells by considering both the data symbol and neighboring cell data, employing a neural network to program memory cells with optimized threshold voltage targets, and using a shared pulses programming scheme to mitigate ICI.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional modulation methods are used that allocate threshold voltage targets based solely on data symbol, then the programming process is simple, but inter-cell interference cannot be fully accounted for leading to higher error rates
Solution Approach 1:
The patent changes the parameter used for threshold voltage target allocation from only data symbol to a combination of data symbol and neighboring cell data symbols. This parameter expansion allows the system to account for inter-cell interference effects while maintaining a systematic approach to programming.
Solution Approach 2:
The patent introduces an intermediary mechanism (neural network or lookup table) that processes both the data symbol and neighboring cell data symbols to determine the optimal threshold voltage target. This intermediary handles the complexity of ICI compensation, separating it from the basic programming logic.
2Duration of action of stationary object
If inter-cell interference is not compensated, then the programming process is fast and simple, but the endurance and data retention time are limited
Solution Approach 1:
The patent applies preliminary action by compensating for inter-cell interference during the programming phase rather than during read operations. By adjusting threshold voltage targets based on neighboring cell data before programming, the system prevents ICI effects from degrading data retention and endurance, rather than correcting them later.
3Reliability
If threshold voltage targets are optimized using AI considering neighboring cells, then endurance is enhanced by 30% in RBER, but the processing complexity increases
Solution Approach 1:
The patent uses a lookup table as a simplified copy of the AI neural network's functionality. Instead of requiring full AI processing during operation, the system pre-computes and stores the mapping between data symbols, neighboring cell patterns, and optimal threshold voltage targets. This allows the system to achieve AI-level ICI compensation with simpler, faster lookup operations.
Data Source
AI summary
Systems, devices, and methods for decoding information bits obtained from storage, including obtaining a plurality of data symbols; providing the plurality of data symbols to a neural network; obtaining a plurality of threshold voltage targets based on an output of the neural network; and programming the plurality of data symbols to a plurality of memory cells included in a storage device based on the plurality of threshold voltage targets.