The strong evolution of systems on chip leads to an emerging trend which is to increase the number of processors on chip. An efficient programming of those multiprocessors needs not only a deep acknowlegment of the architecture itself but also of the algorithms and of the parallel programming model involved in the software. The goal of this course is to train students to develop efficient message passing parallel applications based on standard functionalities of the MPI-1 library.
Parallel Programming for embedded multiprocessors
Last Modification : Monday 14 December 2009