Bin Labeling Schemes for Solid State Storage Cells
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current bin labeling techniques for solid state storage are inefficient, requiring multiple reads and writes, which consume power and can lead to errors in soft read value generation due to misreads being persistently assigned to incorrect bins.
Innovation Solution
Direct write and accumulative bin labeling techniques are employed, where bin labels are updated directly or accumulatively, reducing the need for repeated reads and writes, and mitigating errors by providing better likelihood ratios during soft read value generation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional bin labeling techniques are used with multiple reads, then bin tracking is achieved, but power consumption increases and error rates increase due to repeated reads and writes
Solution Approach 1:
The patent performs bin labeling updates in advance during the read process itself, rather than requiring separate subsequent reads. By updating bin labels immediately when voltage levels are measured, the system eliminates the need for repeated reads to track bin assignments, thereby reducing power consumption while maintaining accurate error tracking for soft read value generation
Solution Approach 2:
The system uses the voltage measurement data obtained during normal read operations to simultaneously update bin labels and track cell states. This self-service approach means the same read data serves dual purposes: determining read values and updating bin assignments, eliminating redundant operations and reducing overall power consumption while improving reliability
2Measurement precision
If multiple reads are performed to update bin labels, then bin tracking accuracy is maintained, but the number of read/write operations increases causing higher power consumption
Solution Approach 1:
The patent combines the bin labeling update operation with the voltage measurement process. By integrating these two functions into a single operational step, the system maintains accurate bin tracking without requiring separate read/write cycles, thereby improving operational efficiency while preserving measurement precision
Solution Approach 2:
The read operation is designed to serve multiple functions simultaneously: it determines the read value of cells, updates bin labels based on voltage levels, and tracks cell state changes. This multi-functional approach eliminates the need for dedicated bin update operations, improving productivity while maintaining bin tracking accuracy
3Reliability
If traditional bin labeling is used, then bin identification is achieved, but misreads are persistently assigned to incorrect bins reducing soft read value accuracy
Solution Approach 1:
The system continuously updates bin labels based on actual voltage measurements and uses this feedback to correct misread assignments. By comparing measured voltage levels against bin thresholds and updating bin labels in real-time, the system can identify and correct erroneous assignments, improving soft read value generation accuracy while preventing persistent misread errors
4Reliability
If more reads are stored with the same number of bits, then noise tolerance improves, but bin label storage capacity is limited
Solution Approach 1:
The patent changes the representation parameters of bin labels from traditional fixed-width binary encoding to a variable-length encoding scheme that utilizes bit patterns more efficiently. This parameter change allows the same number of bits to represent a larger number of bin states, enabling the storage of more read operations and improving noise tolerance without increasing storage capacity
Data Source
AI summary
Information associated with a read to solid state storage is received, including a read number and a read value. The read value is written to a location in a cell and bin map, wherein (1) the location in the cell and bin map corresponds to the read number and (2) the cell and bin map tracks, for each cell in a group of cells, which bin out of a plurality of bins a given cell falls into.