Power density aware application mapping in mesh-based network-on-chip architecture: An evolutionary multi-objective approach

In the era of many-core chips, problem power density is a serious challenge. This particularly important in Network-on-Chip (NoC)-based systems, where application mapping determines resulting patterns and workload distribution across entire chip. Despite this fact, majority algorithms focus on performance, are largely ignored. work investigates problem. Three different pattern metrics with scopes defined, namely, peak, range, regional density. The results using them as objectives together communication cost multi-objective evolutionary approach investigated. Results show that employing results-in Pareto fronts features. analysed discussed. Moreover, case study thermal analysis performed. only large hotspots which translates into higher peak range chip temperatures. proposed shown to significantly improve balancing (up 55%) temperature 7.77%). These indicate importance considering design NoC-based systems their direct impact reliability performance such systems.