IC manufacturing diagnosis based on statistical analysis techniques

During the production of integrated circuits, variations in the production environment can cause significant drops in yield. Since large amounts of data may have to be processed to diagnose the process conditions, the application of computer tools could greatly aid the engineer responsible for this task. In this paper the authors present a methodology for diagnosis, describe the algorithms, and illustrate applications with results from industrial data. More specifically, this paper presents the algorithms for the analysis of intrawafer variability. Measurements are made on many individual devices or circuits across an entire wafer. This information is used as the input to the diagnosis system. The system uses process, device, and circuit simulators to model the fabrication process. The results of the data analysis are lists of faults that may have caused the variability of measured performances. These faults are represented in terms of the inputs to the process simulator