Semiconductor Processing Unit Error Recovery via Data Preservation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing semiconductor devices fail to effectively manage data integrity when an abnormal processing unit is reset, as there is a risk of data being overwritten or lost, and cannot detect adverse effects on normal processing units during the transition.
Innovation Solution
A semiconductor device with multiple processing units, a processing control unit, and a storage unit that writes information about each processing unit's operations, allowing the control unit to stop the faulty unit and determine if a secondary unit can maintain execution based on stored information, ensuring continuous processing and preventing malfunctions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If an abnormal processing unit is reset after detecting an error, then the processing unit can be restored to normal operation, but data may be overwritten or lost during the period from error detection to reset
Solution Approach 1:
The patent applies preliminary action by writing processing information to storage unit 104 before the abnormal processing unit is reset. The writing unit 103 records data about the abnormal processing unit's operations in advance, so that even if data is lost during reset, the processing information is already preserved in the storage unit, preventing information loss while maintaining the ability to restore the processing unit.
2Productivity
If a normal processing unit continues processing after an abnormal processing unit becomes abnormal, then processing continuity is maintained, but the normal processing unit may be adversely affected by the abnormal processing unit's operations
Solution Approach 1:
The patent implements feedback by having the processing control unit 102 monitor processing information stored in the storage unit 104. The control unit checks whether the abnormal processing unit's operations have adversely affected other processing units based on the recorded processing information, and can detect and prevent potential harm while maintaining processing continuity through normal units.
3Reliability
If multiple processing units are placed in a single semiconductor device to enable task takeover, then system reliability is improved, but the complexity of managing data integrity and error detection increases
Solution Approach 1:
The patent introduces an intermediary approach by using a storage unit 104 that acts as a mediator between multiple processing units 101. The writing unit 103 and processing control unit 102 manage the storage unit to record and monitor processing information, simplifying the coordination and error management between processing units while maintaining the ability to detect adverse effects and ensure data integrity.
Data Source
AI summary
A semiconductor device (1) includes a first processing unit (10-1), a second processing unit (10-2), a writing unit (12), a storage unit (14), and a processing control unit (20). The writing unit (12) writes first information related to processing of each of the first processing unit (10-1) and the second processing unit (10-2) into the storage unit (14). The processing control unit (20) controls the operations of the first processing unit (10-1) and the second processing unit (10-2). The processing control unit (20) performs control to stop the first processing unit (10-1) when an error occurs in the first processing unit (10-1). When it is determined that the second processing unit (10-2) where an error has not occurred is able to maintain execution of the first processing by using first information stored in the storage unit (14), the second processing unit (10-2) maintains execution of the first processing.