This program exploits the Contraint Programming feature of OR-Tools (https://code.google.com/p/or-tools/) to implement an optimal solver for the task partitioning problem.

More precisely, given the model of the hardware (number and type of processors and communication links) and the workload (in terms of DAG - see https://github.com/mbambagini/taskdep), the task-to-CPU assignment and scheduling which minimizes the makespan is found.