MPI Detach - Asynchronous Local Completion

Abstract

When aiming for large scale parallel computing, waiting time due to network latency, synchronization, and load imbalance are the primary opponents of high parallel efficiency. A common approach to hide latency with computation is the use of non-blocking communication. In the presence of a consistent load imbalance, synchronization cost is just the visible symptom of the load imbalance. Tasking approaches as in OpenMP, TBB, OmpSs, or C++20 coroutines promise to expose a higher degree of concurrency, which can be distributed on available execution units and significantly increase load balance. Available MPI non-blocking functionality does not integrate seamlessly into such tasking parallelization. In this work, we present a slim extension of the MPI interface to allow seamless integration of non-blocking communication with available concepts of asynchronous execution in OpenMP and C++.

Authors 5

1. Joachim Protze Aachen

   RWTH Aachen University

   Affiliation as printed

   RWTH Aachen University ITC, Germany

2. Marc-André Hermanns Aachen

   RWTH Aachen University

   Affiliation as printed

   RWTH Aachen University ITC, Germany

3. Ali Can Demiralp Aachen

   RWTH Aachen University

   Affiliation as printed

   RWTH Aachen University VCI, Germany

4. Matthias Müller Aachen

   RWTH Aachen University

   Affiliation as printed

   RWTH Aachen University ITC, Germany

5. Torsten Kuhlen Aachen

   RWTH Aachen University

   Affiliation as printed

   RWTH Aachen University VCI, Germany