A hierarchical model to manage hardware topology in MPI applications

The MPI standard is a major contribution in the landscape of parallel programming. Since its inception in the mid 90's it has ensured portability and performance for parallel applications on a wide spectrum of machines and architectures. With the advent of multicore machines, understanding and taking into account the underlying physical topology and memory hierarchy have become of paramount importance. The MPI standard in its current state, however, and despite recent evolutions is still unable to offer mechanisms to achieve this. In this paper, we detail several additions to the standard that give the user tools to address the hardware topology and data locality issues while improving application performance.