For design for manufacturing (DFM) models and methods to be defensible and adoptable, they have to be as simple as possible without losing physical justification. Our ongoing research is re-evaluating several assumptions that CAD research has made in recent years to avoid “DFM overkill”. Starting from a deep understanding of the makeup of different physical phenomena that lead to the observed electrical variability, we have developed variability models and methods in physical design to deal with patterning constraints. <\/p>\n\n\n\n