Detecting Context Sensitive Hotspots in Standard Cell Libraries

In this work, we present a process-aware placement sensitivity analysis flow which systematically inspects a standard cell library and categorizes the potential hot spots induced by optical proximity effect. The hot spots are classified by their susceptibility to context optimization. Hot spots shown to be present in most context environments should be reported back to cell designers and the master cell layout should be fixed. On the other hand, we show that manufacturability-aware cell placement algorithms can significantly reduce the number of hot spots identified as manageable by careful context selection.