Nonvolatile Memory Dummy Wordline Error Correction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Nonvolatile memory devices face challenges in ensuring data reliability and efficient recovery during sudden power-offs, as program operations are often interrupted, leading to difficulties in maintaining data integrity.
Innovation Solution
A method is introduced that differentiates between dummy and normal wordlines by determining the number of memory cells with a specific threshold voltage state, performing error correction only on normal wordlines and skipping it on dummy wordlines, and adjusting verify voltage levels for programming and reading operations to distinguish between user and dummy data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If error correction is performed on all wordlines including dummy wordlines, then data reliability is improved, but recovery time increases and processing efficiency deteriorates
Solution Approach 1:
The patent extracts and identifies dummy wordlines from the set of all wordlines by detecting specific threshold voltage characteristics. Once dummy wordlines are identified, error correction processing is excluded from these wordlines, thereby removing unnecessary processing steps and reducing overall recovery time while maintaining data reliability for normal wordlines.
Solution Approach 2:
The patent applies different processing strategies to different types of wordlines: error correction is applied to normal wordlines but skipped for dummy wordlines. This local differentiation is achieved by detecting the number of memory cells in specific threshold voltage states, allowing the system to adapt processing intensity to the actual data requirements of each wordline type.
2Quantity of substance
If program operation is performed multiple times on each wordline to store multi-bit data, then data capacity is improved, but vulnerability to sudden power-off increases
Solution Approach 1:
The patent performs preliminary detection of the programming status by reading threshold voltage states before attempting error correction. This preliminary action identifies which wordlines have completed programming and which are still in progress, allowing the system to handle incomplete operations appropriately and maintain data integrity even when power is suddenly interrupted.
Solution Approach 2:
The patent uses feedback from threshold voltage readings to determine the programming status of each wordline. By continuously monitoring the number of memory cells in specific threshold voltage states, the system can detect whether programming is complete or interrupted, and adjust subsequent error correction operations accordingly to ensure data reliability.
3Productivity
If verify voltage levels are adjusted to distinguish dummy data from user data, then processing efficiency is improved, but measurement complexity increases
Solution Approach 1:
The patent changes the voltage parameter by applying different verify voltage levels for reading memory cells on dummy wordlines versus normal wordlines. This parameter change enables the system to distinguish between dummy data and user data based on threshold voltage characteristics, improving processing efficiency by allowing selective error correction while using a relatively simple detection mechanism based on voltage state counting.
Data Source
AI summary
A method of operating a memory system includes reading data of first memory cells, the first memory cells being connected to a first wordline from among a plurality of wordlines, the plurality of wordlines including one or more dummy wordlines and one or more normal wordlines; determining whether the first wordline is one of the one or more dummy wordlines by determining, based on the read data, a number of the first memory cells having a first threshold voltage state, the one or more dummy wordlines being wordlines the memory cells of which have been programmed with dummy data, the one or more normal wordlines being wordlines that are not dummy wordlines; and performing a repair algorithm for correcting an error in the read data, selectively according to a result of the determination.