This version of FAµST brings some optimizations in CPU calculation of Faust norms as illustrated by the following figure.
See the API documentation of Faust.norm for further details about these optimisations (especially the full_array and the batch_size arguments).