At advanced process nodes, Liberty or library (.lib) requirements are more demanding due to design complexities, increased number of corners required for timing signoff, and the need for statistical variation modeling. This results in an increase in size, complexity, and the number of .lib characterizations. Validation and verification of these complex and large .lib files is a challenging task and poses a significant threat to successful timing closure and even silicon failures if the .lib errors are not detected and fixed in time.
This white paper describes the use of machine learning (ML) techniques in the Siemens EDA Solido™ Characterization Suite that accelerates production quality .lib characterization and verification at advanced technology nodes. These ML techniques address some of the fundamental challenges with the demanding .lib requirements of modern technology nodes and their validation.