AI-Driven Data Feed Forward: Scaling Advanced Test Methodologies for Chiplet-Based Packages
Speaker: Greg Prewitt, Director of Product Management for Exensio® Test Operations, PDF Solutions
Abstract:
The transition from monolithic dies to chiplet-based architectures in advanced semiconductor packaging leads to a fundamental shift in test complexity. This presentation introduces Data Feed Forward (DFF) infrastructure, designed to collect, transform, transport, and apply test data across the full supply chain, enabling a new generation of data and AI driven test methodologies at scale. At its core, DFF enables test data generated at early production steps to be made available at later test operations, potentially at geographically remote facilities. This unlocks three categories of value: efficiency gains (optimizing test coverage, skipping burn-in for predicted-good devices); quality improvements (predicting trim targets to prevent mis-trims, detecting parameter drift across insertions); and performance enhancements (enriching inline real-time models with historical test data).Central to this infrastructure are Exensio® Test Operations which collects, monitors, controls, and optimizes semiconductor test processes in real time, and Exensio® StudioAI which enables ModelOps for Test, the ability to manage model lifecycle, from training on Exensio’s data repository to deployment. The presentation concludes with a discussion of the broader landscape of advanced package applications, from assembly optimization for grade and performance, through system-level test refinement.
Keywords: Agentic AI, Data Integration, Data Feed Forward (DFF)
Download the presentation or watch the video today!
By downloading this information, I agree that PDF Solutions, Inc. may use my information provided herein to contact me about its products and services, even if I have previously unsubscribed from such communications. For information on our privacy practices and commitment to protecting your privacy, please review our Privacy Policy.