Redundant Die Replacement in Stacked ICs for Defect Repair
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The complexity and cost associated with manufacturing stacked die integrated circuits are increased due to defects, which are difficult and expensive to remedy, as traditional scaling methods reach physical limitations and interconnect constraints.
Innovation Solution
Incorporating a redundant die in the stack that can replace defective dies, either by utilizing a redundant module for repairable defects or a replacement die for non-repairable defects, with a die bus system managing communication and power distribution to ensure seamless functionality preservation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If multiple dies are stacked to improve performance and reduce footprint, then integration density and performance are improved, but the likelihood of defects and difficulty of repair increase
Solution Approach 1:
A redundant die is included in the die stack before operation, prepared in advance as a backup. This preliminary action ensures that if a defect occurs in any functional die, the redundant die can immediately replace it, thereby maintaining system reliability while allowing higher integration density with multiple dies.
Solution Approach 2:
The system changes the operational state of dies in the stack by deactivating defective dies and activating the redundant die. This parameter change (from active to inactive state) allows the system to adapt to defects and maintain functionality, resolving the contradiction between high integration density and defect susceptibility.
2Productivity
If traditional scaling methods are used to improve performance, then manufacturing simplicity is maintained, but physical limitations and power density constraints prevent further improvement
Solution Approach 1:
Instead of continuing to scale individual dies (two-dimensional scaling), the patent transitions to three-dimensional stacking of multiple dies. This dimensional change allows continued performance improvement by adding more functional units vertically, overcoming the physical limitations of traditional planar scaling while maintaining manufacturing feasibility through standardized stacking processes.
3Reliability
If defects occur in stacked die integrated circuits, then system reliability deteriorates, but repair is difficult and expensive
Solution Approach 1:
When a defect is detected in a die within the stack, the defective die is extracted from the active system by deactivating it. The redundant die is then activated to replace the defective one, effectively taking out the harmful element (defective die) from the system without requiring physical removal or complex repair procedures, thus maintaining reliability while simplifying the repair process.
Solution Approach 2:
A redundant die, which is a copy of the functional dies in the stack, is maintained as a backup. When a defect occurs, this copy replaces the defective original, allowing system recovery without complex repair. This copying approach simplifies repair by using an identical replacement unit rather than attempting to fix the defective die.
4Ease of repair
If a redundant die is added to enable replacement, then ease of repair is improved, but device complexity increases
Solution Approach 1:
The redundant die is designed with universal functionality identical to the other dies in the stack, allowing it to perform any function required by the system. This multi-functionality means the redundant die can replace any defective die regardless of its specific role, simplifying the repair process while the uniform design minimizes the increase in overall device complexity.
Solution Approach 2:
The die stack includes self-diagnostic capabilities that automatically detect defective dies and trigger the replacement process by activating the redundant die. This self-service approach reduces the need for external intervention and complex repair procedures, improving ease of repair while the automated nature of the process helps manage device complexity.
Data Source
AI summary
The dies of a stacked die IC are tested and, in response to detection of a defect at one of the dies, the type of defect is identified. If the defect is identified as a defective module repairable at the die itself, a redundant module of the die is used to replace the functionality of the defective module. If the defect is identified as one that is not repairable, a replacement die in the die stack is used to replace the functionality of the defective die.