HEURISTICS FOR DYNAMIC TASK AND COMMUNICATIONS MAPPING IN NoC-BASED HETEROGENEOUS MPSoCs

摘要

To fulfill the need of intensive embedded computations, architects have proposed Network-on-Chip (NoC) based Multi-Processor Systems-on-Chip (MPSoCs). Applications exploit the distinct features of the different types of processors in MPSoCs to optimize the performance metrics: overall execution time, energy consumption, resource usage, etc. Application designers often map statically the tasks on the processors. This static mapping cannot handle many kinds of applications such as those with dynamic workloads and one must use dynamic mapping when several applications run concurrently. The conventional dynamic mapping approaches use simple deterministic routing mechanisms such as XY routing. In this paper, we propose heuristics for dynamic task and communications mapping that considers the placement of tasks and their communications in order to optimize the overall performance. The mapping technique uses a newly proposed Manhattan packing strategy to find the placement of tasks and a newly proposed Multi-Objective Routing Algorithm (MORA) to place communications between the tasks. The placement we propose of the tasks and of their communications leads to a better optimization of several performance metrics (time and energy consumption). Experimental results show that the proposed mapping approach provides significant performance improvements when compared to existing strategies for task dynamic mapping and to those using XY routing.