AI Framework Optimizes X-Ray Inspection for Advanced Chip Packaging
New AI-integrated design guidelines boost defect detection in 2.5D/3D packaging.
Advanced packaging technologies like 2.5D and 3D integration are key to meeting demands for performance and miniaturization, but they introduce metrology challenges due to complex three-dimensional structures. X-ray imaging, essential for non-destructive inspection, often struggles with material density similarities and noise scattering. To address this, a team from academia proposes a framework based on AI-integrated Design of Experiment (DoE) that optimizes X-ray compatibility during the design stage.
The framework is validated through a case study on Chip-on-Wafer-on-Substrate (CoWoS) packaging, a leading advanced packaging technology. By systematically analyzing design parameters and material properties, the AI model predicts outcomes and optimizes processes for high-quality X-ray images. Implementation of these design guidelines promise significantly improved inspection accuracy and reliability, leading to lower production costs and enabling scalable, efficient semiconductor manufacturing.
- AI-integrated Design of Experiment (DoE) framework optimizes X-ray parameters for advanced chip packaging.
- Demonstrated on CoWoS (Chip-on-Wafer-on-Substrate) packaging, a key 2.5D integration technology.
- Promises improved defect detection accuracy and reduced production costs in semiconductor manufacturing.
Why It Matters
Ensures reliable quality control for cutting-edge chip packaging, crucial for scaling AI and HPC hardware.