Guest Editorial: Communication-Aware Designs and Methodologies for Reliable and Adaptable On-Chip AI SubSystems and Accelerators