Core Viewpoint - Yield has long been regarded as the lifeline for chip manufacturers and foundries, being crucial for cost control, production efficiency, capacity increase, and market competitiveness [1] Group 1: Yield Improvement Challenges - The issue of yield in advanced processes has evolved from single-point defects to more systemic, multi-source, and cross-layer complex characteristics due to increased design and process complexity [1] - Traditional DFT tools often struggle to effectively identify complex failure modes caused by manufacturing deviations, layout hotspots, and test boundary effects [1] Group 2: YAD Platform Overview - The Guangliwei YAD yield perception big data diagnostic analysis platform was developed to address the challenges in yield analysis [2] - YAD integrates with the Guangliwei DATAEXP big data analysis platform and the Yield Management System (YMS), enabling real-time correlation of cross-domain data, covering the entire "design-manufacturing-testing-analysis" data diagnostic map [2][4] Group 3: Customer Value and Benefits - YAD enhances analysis efficiency, reducing the analysis time from weeks to hours through a powerful graphical interface and reporting functions [8][9] - It improves root cause analysis accuracy by utilizing AI algorithms for multi-dimensional identification of failure causes and automatically recommending PFA candidates [9] - The platform identifies hidden systemic design issues by incorporating design information into yield analysis [9] - YAD supports multi-dimensional data analysis and verification, dynamically adjusting the analysis scope through deep integration with YMS [9] - The system is highly automated, minimizing manual intervention and enabling systematic data management [9] Group 4: Intelligent Diagnostic Analysis - YAD employs data mining and advanced AI algorithms for accurate root cause analysis (RCA), supporting the identification of specific layout patterns causing systemic failures [11] - The platform generates defect root cause probability maps and intelligently recommends PFA candidates, facilitating quick problem localization [11][15] Group 5: Visualization and Reporting - YAD provides visual analysis of diagnostic reports, including circuit diagrams and layout information, significantly saving time [14] - The platform supports intuitive Wafer Map views for flexible data correlation analysis, aiding efficient diagnostic processes [14] Group 6: Future Directions - Guangliwei YAD transforms dispersed design, process, and testing data into core momentum for yield improvement, breaking down data silos and utilizing AI algorithms to quickly identify key failure modes [16]