Non-Volatile Memory Recovery for SoC Corruption Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Non-volatile memory (NVM) on system on chip (SoC) is susceptible to corruption due to magnetic interference, leading to impaired device functionality, and existing systems lack effective detection and recovery mechanisms.
Innovation Solution
A system on chip with a corruption detection module and a recovery routine that retrieves a recovery image from alternate storage, using secure boot ROM to restore the memory, with mechanisms for integrity checks and external communication interfaces.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If MRAM is used for non-volatile storage on system on chip, then data persistence and functionality are improved, but susceptibility to magnetic interference and corruption increases
Solution Approach 1:
The patent implements preliminary action by storing a recovery image in alternate non-volatile memory before corruption occurs. The corruption detection module continuously monitors memory integrity, and when corruption is detected, the recovery routine automatically retrieves the pre-stored recovery image and restores the corrupted NVM, preventing permanent data loss and maintaining system functionality.
Solution Approach 2:
The patent introduces alternate non-volatile memory as an intermediary storage medium. This alternate memory serves as a backup copy that can be retrieved during recovery operations. The recovery routine acts as an intermediary process that coordinates between the corrupted NVM, the alternate storage, and the processing components to perform seamless restoration.
2Reliability
If recovery mechanisms are added to detect and restore corrupted memory, then system reliability is improved, but device complexity increases
Solution Approach 1:
The patent merges the recovery functionality into the existing boot ROM and processing architecture. The corruption detection module integrates with the secure boot ROM, and the recovery routine is executed through existing processing components. This merging approach allows recovery capabilities to be added without requiring entirely separate hardware systems, thereby limiting the increase in device complexity.
Solution Approach 2:
The system implements self-service through automatic corruption detection and self-repair capabilities. The corruption detection module continuously monitors memory integrity without external intervention, and when corruption is detected, the recovery routine automatically executes to restore the NVM using the alternate storage backup, eliminating the need for manual intervention or external recovery tools.
3Difficulty of detecting and measuring
If continuous monitoring of memory integrity is implemented, then detection capability is improved, but energy consumption increases
Solution Approach 1:
The patent implements periodic action through the corruption detection module that monitors memory integrity at specific intervals rather than continuously. The detection occurs during boot-up, before critical operations, and at scheduled intervals, providing adequate monitoring coverage while minimizing energy consumption compared to continuous real-time monitoring.
Data Source
AI summary
A memory recovery procedure is disclosed for non-volatile memory such as magneto-resistive random access memory for a system on chip. The non-volatile memory stores operating system related code or data. A corruption detection module determines whether the non-volatile memory has been corrupted. A hard coded memory including a recovery routine is operable to access a recovery image stored on an alternate storage system and load the recovery image to the non-volatile memory.