Outlier model using fmax to predict failing devices

In the fabrication and manufacturing of Very Large Scale Integrated Circuits (VLSI) it is needed to continually seek out new methods to reducing the cost associated with testing and insuring reliability. Using the measured maximum operating frequency of a device compared to the population that the device came from outliers can be identified. Analysis of the characteristics of these outliers predict defective units from other populations by measuring the maximum operating frequency.