SSD Read Voltage Calibration Using Random Code Word Sampling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current SSDs face inefficiencies in read voltage calibration, particularly in 3D NAND flash memory, as existing methods are time-consuming and may degrade host I/O data transfer performance, necessitating a more efficient approach to update reference voltage levels across all pages.
Innovation Solution
A method involving a random selection of code words from each page for read voltage calibration, using a random number generator to ensure depth and coverage across all memory cells, allowing for the derivation of updated read voltage levels without evaluating every page, thereby maintaining performance during host operations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional read voltage calibration evaluates every page, then measurement precision is improved, but loss of time increases and productivity decreases
Solution Approach 1:
The patent segments the calibration process by dividing all pages into multiple groups and selecting representative pages from each group. Instead of calibrating every page individually, the system performs calibration on selected representative pages and applies the calibrated read voltages to all pages within each group, thereby reducing calibration time while maintaining acceptable accuracy through group-based segmentation.
Solution Approach 2:
The patent uses copying by calibrating read voltages on selected representative pages and then copying these calibrated voltage settings to all other pages in the same group. This approach assumes that pages within a group have similar characteristics, allowing the calibration results from one page to be replicated across multiple pages, significantly reducing the overall calibration time.
2Measurement precision
If traditional read voltage calibration evaluates every page, then measurement precision is improved, but productivity deteriorates
Solution Approach 1:
The patent segments the calibration workload by selecting only representative pages from each group for actual calibration operations. This segmentation reduces the number of calibration operations from proportional to total page count to a fixed number based on group count, thereby improving productivity and host I/O performance while maintaining sufficient calibration accuracy through the representative sampling approach.
Solution Approach 2:
The patent applies partial action by performing calibration on only a subset of pages (representative pages) rather than all pages. This partial calibration approach is sufficient to maintain read voltage accuracy across all pages while significantly reducing the time and resources consumed, thereby improving overall system productivity and host I/O data transfer performance.
3Loss of time
If random code word selection is used, then loss of time is reduced, but measurement precision may deteriorate
Solution Approach 1:
The patent combines segmentation with random selection by dividing pages into groups and then randomly selecting representative pages from each segmented group. This approach maintains measurement precision by ensuring coverage across all page groups while reducing calibration time through the random selection of only one page per group, rather than calibrating all pages.
Solution Approach 2:
The patent changes the calibration approach parameter from deterministic (evaluating every page) to probabilistic (random selection of representative pages). This parameter change allows the system to achieve sufficient calibration accuracy through statistical sampling while dramatically reducing calibration time, balancing measurement precision with time efficiency.
Data Source
AI summary
Method and apparatus for managing data in a non-volatile memory (NVM) of a storage device, such as a solid-state drive (SSD). In some embodiments, flash memory cells are arranged along word lines to which read voltages are applied to sense programmed states of the memory cells, with the flash memory cells along each word line being configured to concurrently store multiple pages of data. An encoder circuit is configured to apply error correction encoding to input data to form code words having user data bits and code bits, where an integral number of the code words are written to each page. A reference voltage calibration circuit is configured to randomly select a single selected code word from each page and to use the code bits from the single selected code word to generate a set of calibrated read voltages for the associated page.