Dual Calibration Scheme for NVRAM Read Voltage Offset Management
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
As storage capacity increases in non-volatile random access memory (NVRAM), the number of read voltage offset values per page grows, leading to significant metadata storage and calibration overhead, which compromises memory reliability and performance due to increased processing demands.
Innovation Solution
Implementing a dual calibration scheme where a first scheme calculates fewer independent read voltage offset values for nominal operation and a second scheme with more accurate but computationally intensive calculations for error recovery, reducing calibration overhead while ensuring efficient memory performance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If all read voltages are calibrated independently to improve accuracy, then measurement precision improves, but device complexity and processing overhead increase significantly
Solution Approach 1:
The patent segments the calibration process into two distinct schemes: a first calibration scheme for nominal operation that uses fewer independent read voltage offset values, and a second calibration scheme for error recovery that uses more independent read voltage offset values. This segmentation allows the system to use computational resources efficiently by applying the more accurate but computationally intensive second scheme only when needed, while using the lighter first scheme during normal operation.
Solution Approach 2:
The patent implements dynamic calibration strategy where the calibration scheme is selected based on operational conditions. The system dynamically switches between the first and second calibration schemes depending on whether the memory is operating normally or experiencing errors, optimizing the balance between processing overhead and calibration accuracy in real-time.
2Measurement precision
If more independent read voltage offset values are calculated to improve calibration accuracy, then measurement precision improves, but loss of time increases due to extended calibration duration
Solution Approach 1:
The patent divides the calibration process into two segments with different computational intensities. The first calibration scheme performs calibration with fewer independent read voltage offset values, completing the task faster for nominal operation. The second calibration scheme performs more thorough calibration with more independent read voltage offset values, taking longer but providing higher accuracy when needed for error recovery.
Solution Approach 2:
The patent applies partial action by using the first calibration scheme that performs sufficient calibration for normal operation without the full computational intensity of the second scheme. The second scheme is applied excessively (with more independent read voltage offset values) only when error recovery is needed, providing more than enough accuracy for the critical error recovery scenario.
3Reliability
If more read voltage offset values are maintained per page to improve reliability, then reliability improves, but device complexity increases due to increased metadata storage requirements
Solution Approach 1:
The patent segments the storage of read voltage offset values into two categories: common read voltage offset values that are shared across multiple pages, and independent read voltage offset values that are specific to individual pages. This segmentation allows the system to maintain reliability through selective storage of independent offset values only when necessary, while leveraging common offset values to reduce overall metadata storage requirements.
Solution Approach 2:
The patent implements universality by using common read voltage offset values that serve multiple pages simultaneously. These common offset values reduce the total number of unique offset values that need to be stored in metadata, thereby reducing storage overhead while still maintaining the reliability needed for proper read operations across all pages.
Data Source
AI summary
A computer-implemented method, according to one approach, includes: using a first calibration scheme to calibrate the given page in the block by calculating a first number of independent read voltage offset values for the given page. An attempt is made to read the calibrated given page, and in response to determining that an error correction code failure occurred when attempting to read the calibrated given page, a second calibration scheme is used to recalibrate the given page in the block. The second calibration scheme is configured to calculate a second number of independent read voltage offset values for the given page. An attempt to read the recalibrated given page is also made. In response to determining that an error correction code failure did occur when attempting to read the recalibrated given page, one or more instructions to relocate data stored in the given page are sent.