Benefits of MPI Sessions for GPU MPI applications - Archive ouverte HAL Access content directly
Conference Papers Year :

Benefits of MPI Sessions for GPU MPI applications

(1) , (2, 3) , (2, 3) , (2, 3) , (1) , (1)
1
2
3

Abstract

Heterogeneous supercomputers are now considered the most valuable solution to reach the Exascale. Nowadays, we can frequently observe that compute nodes are composed of more than one GPU accelerator. Programming such architectures efficiently is challenging. MPI is the defacto standard for distributed computing. CUDAaware libraries were introduced to ease GPU inter-nodes communications. However, they induce some overhead that can degrade overall performances. MPI 4.0 Specification draft introduces the MPI Sessions model which offers the ability to initialize specific resources for a specific component of the application. In this paper, we present a way to reduce the overhead induced by CUDA-aware libraries with a solution inspired by MPI Sessions. In this way, we minimize the overhead induced by GPUs in an MPI context and allow to improve CPU + GPU programs efficiency. We evaluate our approach on various micro-benchmarks and some proxy applications like Lulesh, MiniFE, Quicksilver, and Cloverleaf. We demonstrate how this approach can provide up to a 7x speedup compared to the standard MPI model.
Fichier principal
Vignette du fichier
gpu_mpi_sessions.pdf (2.42 Mo) Télécharger le fichier
Origin : Files produced by the author(s)

Dates and versions

cea-03322976 , version 1 (20-08-2021)

Identifiers

  • HAL Id : cea-03322976 , version 1

Cite

Maxim Moraru, Adrien Roussel, Marc Pérache, Hugo Taboada, Christophe Jaillet, et al.. Benefits of MPI Sessions for GPU MPI applications. EuroMPI '21 - 28th European MPI Users' Group Meeting, Sep 2021, Leibniz, Germany. ⟨cea-03322976⟩
273 View
307 Download

Share

Gmail Facebook Twitter LinkedIn More