Read threshold estimation system using polynomial regression based calculations
By optimizing the read threshold voltage using a polynomial regression calculation method, the problem of high read error rate in memory systems was solved, resulting in more efficient data reading and improved reliability of memory systems.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- SK HYNIX INC
- Filing Date
- 2022-11-28
- Publication Date
- 2026-07-10
AI Technical Summary
Existing technologies have problems with high read error rates and inability to effectively adjust them when determining the read threshold voltage in a memory system. This is especially true in multi-layer cell memories, where distortion and overlap of threshold voltage distribution lead to read failures.
A calculation method based on multinomial regression is adopted. Through mathematical modeling and multiple calculations, the optimal reading threshold voltage for the next reading operation is estimated. The selection of the reading threshold is optimized by using the product and normalization operation between the input and output matrices.
It improves the accuracy and efficiency of read operations, reduces read errors, and enhances the reliability and data access performance of the memory system.
Smart Images

Figure CN116230031B_ABST