slide 26

A processing architecture for LMNL