The UTC Graduate School is pleased to announce that Evan Suggs will present Master’s research titled, Kokkos-Enhanced ExaMPI on 03/23/2023 at 2:00 PM in Multidisciplinary Research Building Conference Room. Everyone is invited to attend.
Computer Science
Chair: Anthony Skjellum
Co-Chair:
Abstract:
Kokkos provides in-memory advanced data structures, concurrency, and algorithm to support advanced C++ parallel programming. MPI provides the most widely used message passing model for inter-node communication. Many programmers use both Kokkos and MPI together. The purpose of this position paper using Kokkos within an MPI implementation to obtain performance and functionality benefits both for the MPI itself, and for applications that use both Kokkos+MPI. For instance, it will be possible, in this model to pass first-class Kokkos objects directly to extended C++-based MPI APIs. Here, we describe on-going efforts to achieve this type of integrated model using ExaMPI, our C++17-based subset implementation of MPI-4. Working with C++-friendly APIs, and Kokkos extensions, examples of the benefits of functionality and performance are shown. We explain why direct use of Kokkos within the certain parts of the MPI implementation are crucial to getting added performance in addition to expressivity. We will also motivate why making Kokkos memory spaces visible to the MPI implementation generalizes the idea of “CPU memory” and “GPU memory” in ways that provide for further optimizations in heterogeneous Exascale architectures. Besides showing the current state of the prototype, we describe future goals, and how these mesh both with a possible future C++ API for MPI-5 as well as the potential for accelerating MPI on architectures that incorporate accelerators.