Memory Controller Data Identification for Power Failure Recovery
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing memory systems face challenges in reliably restoring data after an unexpected shutdown, as they rely on backup power to write data to non-volatile memory, but the limited capacity of backup power supplies restricts the amount of data that can be rendered non-volatile, leading to potential data loss and inefficiencies in data restoration.
Innovation Solution
A memory system configuration that includes a memory controller and non-volatile memory with a backup power supply, where the controller writes data identification information to the non-volatile memory during backup power supply operation, allowing for data restoration upon main power recovery, and uses a look-up table to manage user data storage and identification, enabling efficient data protection and restoration even with limited backup power capacity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If data is written to non-volatile memory using backup power supply, then data loss is prevented, but the limited backup power capacity restricts the amount of data that can be protected
Solution Approach 1:
The data protection function is segmented into two parts: (1) writing data identification information (such as hash values or metadata) to non-volatile memory during backup power operation, and (2) restoring actual data from volatile memory after main power recovery. This segmentation allows protection of data identification with limited backup power while enabling restoration of much larger data sets afterward.
Solution Approach 2:
The system performs preliminary actions by writing data identification information to non-volatile memory during the backup power supply operation before main power failure occurs. This preliminary recording of data identifiers enables subsequent data restoration without requiring the backup power to sustain the entire data writing process.
2Reliability
If all user data is written to non-volatile memory during backup power operation, then complete data protection is achieved, but the time and energy required exceed backup power capacity
Solution Approach 1:
The patent extracts only the essential data identification information (such as hash values, metadata, or pointers) from the complete user data set and writes this extracted information to non-volatile memory during backup power operation. The actual user data restoration occurs afterward from volatile memory, dramatically reducing the writing time and power requirements during the critical backup period.
3Reliability
If a large backup power supply is used to protect all data, then complete data protection is possible, but the device size and cost increase
Solution Approach 1:
The data protection approach is segmented so that only data identification information needs to be preserved in non-volatile memory during power failure, rather than all user data. This segmentation enables the use of a compact backup power supply that can handle minimal writing operations, while the main data restoration occurs after power recovery without requiring extensive backup capacity.
Data Source
AI summary
According to on embodiment, a memory system includes a non-volatile memory, a memory controller including a data buffer configured to store first data, and a backup power supply configured to supply first power to the non-volatile memory and the memory controller, when second power from an external main power supply is off. The memory controller is configured to write second data including information identifying the first data in the non-volatile memory without writing the first data in the non-volatile memory, when the first power is supplied from the backup power supply, and transmit the information stored in the non-volatile memory to an external host device, after a recovery of the second power from the main power supply.