NAND Flash Read Threshold Modeling for Random Telegraph Noise
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Random telegraph noise (RTN) in NAND flash memory devices causes read errors and device instabilities, hindering scaling and accuracy in multi-level memory cells.
Innovation Solution
Implementing a storage device with a storage controller that performs multiple read operations on a target memory cell at different times to obtain target voltages, selects a threshold model, and generates data using these voltages, thereby reducing the effects of RTN.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple read operations are performed on a memory cell to reduce RTN effects, then read accuracy is improved, but read time increases
Solution Approach 1:
The system performs preliminary characterizations of the memory cell to create a profile that predicts RTN behavior. This preliminary action allows the system to determine the optimal number of read operations needed, avoiding unnecessary reads while ensuring accurate data recovery even in the presence of RTN.
Solution Approach 2:
The system dynamically adjusts the number of read operations based on the memory cell's RTN profile and the specific data being read. Rather than using a fixed number of reads, the system adapts the read count to the individual cell's characteristics and the complexity of the data, optimizing the balance between accuracy and time.
2Manufacturing precision
If multiple threshold models are used to handle different memory cell characteristics, then manufacturing precision is improved, but device complexity increases
Solution Approach 1:
The system performs preliminary characterizations of each memory cell to create an RTN profile during manufacturing or initial operation. This preliminary action enables the system to select the appropriate threshold model for each cell without requiring complex real-time analysis, simplifying the controller's operation while maintaining high precision.
Solution Approach 2:
Instead of using a single universal threshold model for all memory cells, the system assigns different threshold models to different cells based on their individual RTN profiles and characteristics. This local quality approach allows each cell to be handled with the most suitable model, improving overall precision without requiring the controller to manage all possible models simultaneously.
Data Source
AI summary
A storage device including: a non-volatile memory comprising a plurality of memory cells, wherein the plurality of memory cells comprises a target memory cell; and a storage controller: wherein the storage controller is configured to: read the target memory cell at a plurality of target read times to obtain a plurality of target voltages, select a threshold model corresponding to the target memory cell from among a plurality of threshold models, and generate data corresponding to the target memory cell by providing the plurality of target voltages to the threshold model.