Hybrid Approach Emerges For Edge/Cloud Inspection Of Chips

The article discusses the emerging hybrid approach for inspecting chips using a combination of edge and cloud computing. Chipmakers and equipment vendors are facing challenges balancing the need for massive storage and compute resources of the cloud with the faster response time of the edge for making adjustments at the tool level. This approach requires refined machine learning models and access to both upstream and downstream data. Machine learning is being integrated into inspection and metrology processes, with a focus on utilizing AI/ML models for high-speed inspection and metrology purposes. The article emphasizes the importance of combining inspection data with other equipment data throughout the manufacturing process to improve defect detection and speed up root-cause analysis.