Decentralized Defect Scanning Using Component Vintage Hashes
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing technologies for detecting defects in deployed computing systems do not consider the vintage of components, leading to inefficient defect scanning, unnecessary testing, and increased total cost of ownership for cloud service providers, as they lack the capability to scale defect detection across heterogeneous IP blocks from multiple vendors while preserving confidentiality and anonymity.
Innovation Solution
A method and system for decentralized tracking of Defects Per Million (DPM) statistics that considers component vintage, using a hashed ID to identify components and apply customized remedial actions, integrating with Scan-At-Field test patterns to detect defects while preserving IP confidentiality and anonymity, and leveraging a baseboard management controller (BMC) for decentralized defect scanning.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If defect scanning is performed without considering component vintage, then all components are tested uniformly, but this leads to unnecessary testing of low-risk components and increased total cost of ownership
Solution Approach 1:
The patent applies local quality by differentiating defect scanning strategies based on component vintage characteristics. Components are categorized into different risk groups (e.g., high-risk vintages with known defects vs. low-risk vintages), and scanning intensity is adjusted locally for each group. This allows focused testing on vulnerable components while reducing or skipping tests on stable components, thereby improving testing efficiency without compromising defect detection accuracy for critical components.
Solution Approach 2:
The patent changes the parameter of scanning intensity based on vintage-related parameters such as component age, manufacturing batch, and known defect patterns. By dynamically adjusting scanning parameters (e.g., scan frequency, test depth, resource allocation) according to vintage risk profiles, the system achieves both high reliability in defect detection and improved productivity by avoiding unnecessary comprehensive scans of low-risk components.
2Reliability
If comprehensive defect scanning is performed across all components, then defect detection coverage is maximized, but system downtime and operational disruption increase
Solution Approach 1:
The patent implements preliminary action by proactively identifying and prioritizing components with known vintage-related defects before they cause system failures. Using historical defect data and vintage information, the system pre-schedules scanning for high-risk components during maintenance windows or low-utilization periods, while deferring or skipping scans for low-risk components. This allows comprehensive defect detection coverage for critical components without causing unnecessary system downtime.
Solution Approach 2:
The patent applies partial action by performing comprehensive defect scanning only on components with high vintage-related risk profiles, while applying reduced or no scanning to low-risk components. This selective approach maintains high defect detection coverage for vulnerable components while minimizing the time loss and operational disruption associated with scanning all components uniformly.
3Measurement precision
If material vintage information is collected and stored for all components, then defect correlation accuracy is improved, but data storage requirements and processing complexity increase
Solution Approach 1:
The patent extracts and stores only the essential vintage-related parameters needed for defect correlation, such as component age, manufacturing batch ID, and key process variations, rather than collecting all possible component data. This selective extraction maintains high defect correlation accuracy by focusing on the most relevant vintage characteristics while significantly reducing data storage requirements and processing complexity compared to comprehensive data collection.
Solution Approach 2:
The patent segments vintage information into hierarchical categories (e.g., broad manufacturing era, specific batch, sub-batch variations) and stores them in a structured format that enables efficient querying and correlation. This segmentation allows the system to achieve high defect correlation accuracy by organizing data at appropriate levels of granularity without managing excessive detail, thereby reducing data management complexity while maintaining measurement precision.
Data Source
AI summary
Managing scan detection of a component in a computing system includes detecting a scan interrupt, reading a scan register of the component, the scan register including a hashed identifier (ID) of the component; getting material vintage information of the component based at least in part on the hashed ID; and initiating a scan of the component based at least in part on the material vintage information to detect any defects in the component.