Multi-pass Memory Programming Suppresses Read Noise
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Non-volatile memory devices face challenges in accurately programming memory cells due to read noise, which can lead to errors and widening of threshold voltage distributions, especially as devices are scaled down, affecting the reliability of data storage.
Innovation Solution
The solution involves distinguishing noisy memory cells from non-noisy cells during programming by performing additional verify tests and subjecting noisy cells to additional programming, such as soft programming, to ensure the threshold voltage remains above the verify level, thereby reducing read errors and maintaining narrower threshold voltage distributions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If additional verify tests and soft programming are performed on noisy cells, then read error rate is reduced and threshold voltage distribution is narrowed, but programming time and process complexity increase
Solution Approach 1:
The patent performs preliminary verify tests during the programming process to identify noisy cells before final programming is complete. By detecting noisy cells in advance and applying additional programming only to those specific cells, the system prevents read errors without requiring all cells to undergo extended programming, thus reducing overall time loss.
Solution Approach 2:
The patent applies different programming strategies to different cells based on their noise characteristics. Noisy cells receive additional verify tests and soft programming, while non-noisy cells follow the standard programming flow. This localized approach ensures high reliability for problematic cells without unnecessarily extending programming time for all cells.
2Manufacturing precision
If additional verify tests and soft programming are performed on noisy cells, then threshold voltage distribution is narrowed, but device complexity increases
Solution Approach 1:
The system performs preliminary identification of noisy cells through verify tests before implementing the full soft programming sequence. This staged approach simplifies the overall process by separating the complexity of noise detection from the complexity of corrective programming, making the system more manageable.
Solution Approach 2:
The patent implements feedback mechanisms where verify test results are used to determine whether additional programming is needed. This feedback loop allows the system to adapt its behavior based on real-time measurements, reducing complexity by avoiding unnecessary programming steps for cells that pass initial verification.
3Reliability
If multi-pass programming is used to address read noise, then data integrity is improved, but programming time increases
Solution Approach 1:
The patent applies multi-pass programming selectively only to noisy cells that fail verify tests, while non-noisy cells complete programming in a single pass. This localized multi-pass approach maintains high data integrity for problematic cells without significantly reducing overall programming throughput.
Solution Approach 2:
By performing verify tests during the programming process rather than after, the system identifies noisy cells early and applies additional programming only when necessary. This preliminary detection approach ensures data integrity while minimizing the impact on overall programming speed.
Data Source
AI summary
Memory cells which have read noise are identified during a programming pass and an amount of programming is increased for noisy memory cells compared to non-noisy cells. The read noise is indicated by a decrease in the threshold voltage of a cell when the cell is repeatedly read. In one approach, during the programming pass, a cell enters a temporary lockout state when it passes a first verify test and is subject to one or more additional verify tests. Data is stored to identify the cell as a noisy cell or a non-noisy cell based on the one or more additional verify tests. Or, the cells are subject to the one or more additional verify tests at the end of the programming pass. In a subsequent programming pass, the noisy cell is programmed using a stricter verify condition. Or, the noisy cell is kept in an erased state.