Artificial Intelligence (AI) on CMP Edge Residual and Pin Hole Defects

Most automated inspection tools utilize a die-to-die inspection strategy. EAGLEview inspects the entire wafer surface, including partial die near the wafer edge. This allows the EAGLEview to detect subtle defects such as CMP residual Tungsten very close to the wafer edge.  The capability provides an “early-warning” for CMP defects allowing problems to be corrected before they become severe and impact yield.  EAGLEview uses deep learning and artificial intelligence (AI) to automate detection and classification of CMP residual Tungsten defects.