Calibre SONR, a feature-vector driven machine-learning platform, enables a broad range fab solutions and customizable, scalable and integrated design and fab interaction analytics such as care area generation, layout clustering and comparison. It is compatible with the full Calibre eco-system.
Calibre SONR offers a unique clustering method for unsupervised or semi-supervised machine-learning models. Compared to the open source or other commercial clustering or grouping methods, Calibre SONR shows significant performance benefits for the full chip data level (billions) and maintains high accuracy.
Calibre SONR converts all kinds of information to features through its powerful and flexible feature collection platform. Information such as geometry dimension, pattern density, lithography image intensity, as well as customized properties generated from other Calibre tools can all be converted into features with excellent scalability and memory consumption.
Calibre SONR uses a unique database format (MLDB) for data saving and querying. Compared to the traditional CSV and SQL saving, MLDB file size can be compressed to tens or even hundreds of times smaller by using the optimized algorithm. Querying information from the database is very fast (seconds) and can use multiple threads for further speed up.
We help you adopt, deploy, customize and optimize your complex design environments. Direct access to engineering and product development lets us tap into deep domain and subject matter expertise.
The Siemens Support Center provides you with everything in one easy-to-use location -
knowledgebase, product updates, documentation, support cases, license/order information and more.
Across all process nodes and design styles, the Calibre toolsuite delivers accurate, efficient, comprehensive IC verification and optimization, while minimizing resource usage and tapeout schedules.