Processor Integrity Checking With Approximate Calculation Comparison
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for detecting intermittent faults in processing circuitry, such as mercurial cores, are inefficient and costly, particularly when redundant execution or duplicated computation logic is used, and there is a need for a more effective and resource-efficient method to detect calculation errors.
Innovation Solution
Incorporating approximation circuitry to calculate an approximate result of a calculation concurrently with the main calculation circuitry, and using a deviation threshold to compare the results, allowing for efficient integrity checking without duplicating computation resources.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If redundant execution or duplicated computation logic is used to detect intermittent faults, then detection capability is improved, but resource overhead and device complexity increase
Solution Approach 1:
The patent creates a simplified copy of the computation logic in the form of an approximation circuit that generates approximate results. Instead of duplicating the full computation logic, only an approximate version is implemented, which reduces the complexity of the copied structure while maintaining its fault-detection function. This allows the system to detect intermittent faults without the full resource overhead of complete redundancy.
Solution Approach 2:
The patent applies different quality levels to different parts of the computation system. The main computation circuitry operates at full precision, while the approximation circuitry operates at reduced precision. This local differentiation allows the system to maintain high accuracy where needed while using lower-resource approximation for verification purposes, thereby reducing overall device complexity while preserving detection capability.
2Reliability
If redundant execution is used to detect intermittent faults, then detection capability is improved, but processing time increases
Solution Approach 1:
The patent performs only a partial verification by comparing approximate results rather than full re-execution. The approximation circuit generates results quickly without completing the full computation, which is sufficient for detecting faults. This partial action approach enables fault detection without the time penalty of complete redundant execution, as the approximation can be generated and compared much faster than full re-computation.
Solution Approach 2:
The verification process is segmented into two independent parallel paths: the main computation path and the approximation path. This segmentation allows both paths to operate simultaneously without sequential dependency, eliminating the time overhead that would result from executing verification after the main computation. The segmented architecture enables real-time fault detection without delaying the primary computation workflow.
3Device complexity
If approximation circuitry is used for integrity checking, then resource overhead is reduced, but measurement precision decreases
Solution Approach 1:
The patent substitutes the mechanical approach of exact computation with an approximation-based verification mechanism. Instead of using the same precise computation method for both main execution and verification, the system replaces the verification computation with an approximation process. This substitution reduces resource overhead significantly while maintaining sufficient precision for fault detection, as the approximation is only required to detect deviations, not to provide exact results.
Data Source
Figure 1
Figure 2
Figure 3~5
AI summary
An apparatus has processing circuitry to execute instructions. The processing circuitry has calculation circuitry which is responsive to one or more instructions requiring a calculation to be performed to compute the result of the calculation and approximation circuitry which is responsive to said one or more instructions to calculate an approximate result of the calculation independently of the calculation circuitry. The processing circuitry also has integrity checking circuitry to perform an integrity check by comparing the result of the calculation performed by the calculation circuitry and the approximate result of the calculation performed by the approximation circuity. The integrity checking circuitry detects an error in the processing circuitry if it is determined that a difference between the result of the calculation and the approximate result of the calculation is greater than a deviation threshold.