This publication presents a high-impact collaboration between NVIDIA and NenoVision, showcasing how AFM-in-SEM technology fundamentally extends the capabilities of advanced semiconductor failure analysis. In the paper, Chuan Zhang, Lead Failure Analysis Engineer at NVIDIA, provides a clear and practice-driven explanation of why conventional SEM-based techniques often fall short at advanced technology nodes—and how in-situ Conductive AFM (CAFM) overcomes these limitations.
By integrating CAFM directly into a plasma FIB/SEM environment, the demonstrated workflow enables precise electrical defect localization without repeated vacuum breaks, sample transfers, or back-contact preparation. This is particularly critical for modern SOI-based technologies, where traditional back-contacting is no longer feasible. The study highlights real-world case analyses where defects invisible to standard SEM methods are successfully identified using CAFM, including complex stacked-die structures and nanoscale gate failures.
Beyond workflow efficiency, the publication introduces novel capabilities such as Scanning EBIC Microscopy, opening new dimensions for junction-level electrical characterization. Overall, this work provides valuable insight into how AFM-in-SEM solutions are becoming an essential tool for next-generation semiconductor failure analysis, validated through direct industrial application at NVIDIA.
