Non-Volatile Storage Access Performance Optimization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The access performance of non-volatile storage devices deteriorates when programming data of a size smaller than the fixed minimum program number (FMPN), as existing methods require combining such data with additional data to reach the FMPN, leading to inefficient data transfer and reduced performance.
Innovation Solution
Divide the FMPN into three parts: a L side, an M side, and a dummy side, where the L side stores the data to be updated, the M side is read into the storage buffer, and the dummy side is used to combine with the L and M sides to form a final data block that meets the FMPN, optimizing data transfer by minimizing unnecessary data movement.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If data smaller than FMPN is combined with additional data to reach FMPN, then the programming requirement is met, but the data transfer amount increases and access performance deteriorates
Solution Approach 1:
The FMPN data block is segmented into three parts: L side (data to be updated), M side (data to be read), and dummy side (padding data). This segmentation allows selective processing where only necessary data portions are transferred, avoiding the need to transfer entire FMPN blocks when only small updates are required.
Solution Approach 2:
Different portions of the FMPN block are treated differently: the L side receives new data, the M side is read from storage, and the dummy side is filled with padding. This local differentiation allows the system to optimize data transfer by only moving necessary data portions rather than treating the entire block uniformly.
2Quantity of substance
If FMPN is increased, then the programming capacity per operation increases, but the access performance for small data sizes deteriorates
Solution Approach 1:
By segmenting the FMPN block into L side, M side, and dummy side, the system can handle small data updates efficiently even when FMPN is large. Only the necessary L side and M side portions are transferred, while the dummy side is locally generated, preventing performance degradation with larger FMPN values.
Solution Approach 2:
The system performs partial action by only transferring the necessary data portions (L side and M side) rather than the entire FMPN block. The dummy side is handled locally without transfer, representing an excessive action that is avoided through selective processing.
3Reliability
If complement data is read out to maintain validity, then data integrity is preserved, but the data transfer amount increases
Solution Approach 1:
The M side data is selectively read from storage to maintain integrity of existing valid data, while the L side is updated with new data. The dummy side is locally generated without requiring transfer, optimizing the balance between data integrity and transfer efficiency.
Data Source
AI summary
A method for improving access performance of a non-volatile storage device when programming data of a size smaller than a fixed minimum program number (FMPN) is disclosed. The method includes the steps of: predetermining a size of a blank data section for combining with a first data section and a second data section, the total size of the first data section, the second data section and the blank data section equals the FMPN; reading out data located at the second data section; updating a new data to the first data section; combining the new data with the data at the second data section; and incorporating the combined data with the blank data of the blank data section to become a final data, and programming the final data.