Analyzing Data Intensive Networks on Chips

Abstract

A novel framework [2] [3] is proposed to find efficient data-intensive flow distributions on Networks on Chip (NoC). In [3], the authors discussed the virtual cut-through switching, and we extend the flow matrix and analysis in a new switching mechanism, a modified store-and-forward switching mechanism. We explore the various workload distribution applications compared to the previous data load evenly distribution scenario. The new algorithms lead to an efficient makespan and a significant saving in the number of cores used.