Redundant Reconfigurable Clusters for Transparent Defect Substitution
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Reconfigurable circuits face manufacturing yield and premature partial operational failure issues due to their large size, leading to high costs in bypassing defective elements and burn-in processes that do not guarantee long-term reliability.
Innovation Solution
The implementation of redundant reconfigurable clusters within a macro-cluster, where a subset of clusters can be externalized for application usage, with a redundant cluster operating transparently to replace defective ones, allowing configuration data streams to remain unchanged and enabling transparent defect substitution without modifying the configuration data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If reconfigurable circuits are made larger to implement more functions, then functionality and adaptability are improved, but manufacturing yield and reliability deteriorate due to increased defects
Solution Approach 1:
The reconfigurable circuit is divided into multiple independent reconfigurable clusters, each capable of being individually replaced. This segmentation allows defective clusters to be isolated and substituted without affecting the entire circuit, thereby maintaining high functionality while improving manufacturing yield through localized defect management.
Solution Approach 2:
Redundant reconfigurable clusters are built into the circuit architecture in advance, serving as pre-prepared replacements for potentially defective clusters. This beforehand cushioning ensures that if manufacturing defects occur, functional redundancy is already in place to maintain circuit operation, thus improving reliability without sacrificing adaptability.
2Reliability
If laser trimming is used to bypass defective elements, then manufacturing yield is improved, but fabrication cost increases significantly
Solution Approach 1:
The redundant clusters and replacement logic are built into the circuit during standard fabrication processes, performing the defect mitigation action in advance rather than requiring post-fabrication laser trimming. This preliminary action integrates reliability improvement into the manufacturing flow itself, avoiding the need for expensive and time-consuming laser trimming operations.
3Reliability
If burn-in process is used to address premature failures, then reliability is improved, but fabrication cost and processing time increase
Solution Approach 1:
Redundant clusters are built in during standard fabrication, providing immediate protection against premature failures without requiring extended burn-in processing. The replacement capability is pre-established, allowing the circuit to withstand premature failures through structural redundancy rather than time-consuming thermal stress testing.
4Reliability
If redundant clusters are added to replace defective ones, then reliability is improved, but device complexity increases
Solution Approach 1:
The redundant reconfigurable clusters are designed with the same universal interface and configuration protocol as the functional clusters. This universality allows the replacement logic to be simple and uniform - any cluster can replace any other cluster through standard configuration routines, minimizing the increase in device complexity while maximizing defect tolerance.
Data Source
AI summary
A reconfigurable circuit having redundant reconfigurable clusters is described herein.