Calibre LFD and Calibre ML-LFD
Improving design robustness, reliability and yield
Calibre® LFD and Calibre ML-LFD are critical tools within the Calibre Design for Manufacturing (DFM) family designed to enhance IC manufacturability and yield. Calibre LFD supports litho-friendly design by predicting design response to the lithographic process window using process variation (PV) bands. It identifies and scores potential manufacturing failures, guiding designers to create robust IC layouts, reducing process variation.
Calibre ML-LFD employs a supervised machine-learning flow that uses a regression algorithm to locate and predict potential hotspots in layouts. This machine-learning approach, which leverages Deep Neural Network (DNN), enables ML-LFD to achieve high accuracy and bypass the need for computationally-intensive OPC models and recipes.