Flash Memory Data Shaping for Wear Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Flash memory wear and data retention issues arise from repeated storage, leading to reduced endurance and performance due to threshold voltage changes causing noise and cross-coupling effects, especially when programming cells to high threshold voltages.
Innovation Solution
Implementing data shaping techniques, such as the greedy shaping algorithm and biased random number generators, to manipulate data bits and control threshold voltages, reducing the number of cells transitioning from low to high states, thereby minimizing current through the oxide isolation layer and reducing wear.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If data is repeatedly stored in flash memory, then data capacity is increased, but memory wear increases and data retention capability deteriorates
Solution Approach 1:
The patent applies parameter changes by modifying the distribution of data values before storage. The shaping algorithm transforms input data to control the proportion of logical ones and zeros, thereby controlling the distribution of threshold voltages in the memory cells. This parameter transformation reduces the frequency of high-threshold-voltage programming, which in turn reduces memory wear and improves data retention capability while maintaining data capacity.
2Quantity of substance
If memory cells are programmed to high threshold voltage, then data storage capability is improved, but memory wear increases and endurance deteriorates
Solution Approach 1:
The shaping algorithm changes the parameter distribution of data values to reduce the proportion of high-threshold-voltage programming. By controlling the ratio of logical ones to zeros, the algorithm ensures that memory cells are programmed to high threshold voltages less frequently, thereby improving memory endurance while maintaining adequate data storage capability through optimized data distribution.
3Manufacturing precision
If all input data is shaped using the same shaping algorithm, then data distribution control is improved, but variance of programmed data increases
Solution Approach 1:
The patent implements dynamics by making the shaping algorithm adaptive rather than static. The algorithm dynamically adjusts its behavior based on the characteristics of the input data, selecting different shaping strategies or parameters to optimize the output distribution. This dynamic adaptation reduces the variance of programmed data while maintaining precise control over data distribution, resolving the contradiction between control precision and stability.
Data Source
AI summary
A device includes a non-volatile memory and a controller coupled to the non-volatile memory. The controller is configured to determine a first shaping level corresponding to applying a first shaping operation to data to be stored to the non-volatile memory. The controller is further configured to, in response to the first shaping level exceeding a threshold, perform a second shaping operation to generate shaped data that corresponds to the data, the shaped data having a second shaping level that is less than the threshold.