Flash Controller RNN for Dynamic TVSO Read Error Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Characterizing flash memory devices is expensive and time-consuming, and their characteristics change over time, making accurate characterization challenging, especially for Multi-level, Triple-level, and Quad-level cells that require multiple threshold voltage reads to maintain low Uncorrectable Bit Error Rates (UBER).
Innovation Solution
A method involving a Regression Neural Network (RNN) inference model stored on a flash controller to identify Threshold-Voltage-Shift Read-Error (TVS-RE) curves, which monitors usage characteristics like program and erase cycles to determine optimal Threshold-Voltage-Shift Offset (TVSO) values for reduced read errors, ensuring UBER remains within acceptable levels.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If threshold-voltage-shift reads are performed with fixed TVSO values, then read operations are simple and fast, but UBER increases over time as flash memory devices degrade
Solution Approach 1:
The patent applies dynamics by making the TVSO values dynamic rather than fixed. The system adjusts TVSO values based on the current state of the flash memory device, including wear level, temperature, and usage patterns. This allows the read operation to adapt to changing conditions, maintaining low UBER throughout the device lifetime while managing the complexity through adaptive algorithms.
Solution Approach 2:
The patent implements feedback mechanisms where read error rates are monitored and used to adjust TVSO values for subsequent reads. The system continuously learns from actual device behavior and characterization data, refining its TVSO selection strategy to optimize reliability while balancing operational complexity.
2Reliability
If comprehensive device characterization is performed to identify optimal TVSO values, then UBER is maintained at low levels, but the process becomes expensive and time-consuming
Solution Approach 1:
The patent applies preliminary action by performing device characterization during manufacturing and initial device setup, before the device enters normal operation. This advance characterization establishes baseline parameters and optimal TVSO values that are stored for use during normal operation, eliminating the need for time-consuming characterization during production or field use.
Solution Approach 2:
The system implements self-service by automatically performing ongoing characterization and TVSO optimization using the device's own operational data. The flash memory device monitors its own wear, error patterns, and performance metrics, then autonomously adjusts its read parameters without requiring external characterization equipment or manual intervention.
3Measurement precision
If device characteristics are monitored continuously to adjust TVSO values, then read accuracy is maintained, but system complexity increases
Solution Approach 1:
The patent extracts and separates the complex monitoring and characterization functions into dedicated hardware modules and algorithms, distinct from the main flash memory array and control logic. This modular approach isolates the complexity to specific components that can be optimized independently, allowing accurate threshold voltage measurement and TVSO adjustment without unnecessarily complicating the entire flash memory system.
Data Source
AI summary
A method and apparatus for reading a flash memory device are disclosed. A Regression Neural Network (RNN) inference model is stored on a flash controller. The RNN inference model is configured for identifying at least one Threshold-Voltage-Shift Read-Error (TVS-RE) curve that identifies a number of errors as a function of Threshold Voltage Shift Offset (TVSO) values. The operation of a flash memory device is monitored to identify usage characteristic values. A neural network operation of the RNN inference model is performed to generate a TVS-RE curve corresponding to the usage characteristic values. The input for the neural network operation includes the usage characteristic values. A TVSO value is identified corresponding to a minimum value of the TVS-RE curve. A read of the flash memory device is performed using a threshold-voltage-shift read at the TVSO value.