RMA-MT: A Benchmark Suite for Assessing MPI Multi-threaded RMA Performance
Authors:
- Matthew G. F. DosanjhMDMatthew G. F. Dosanjh
- Taylor GrovesTGTaylor Groves
- Ryan E. GrantRGRyan E. Grant
- Ron BrightwellRBRon Brightwell
- Patrick G. BridgesPBPatrick G. Bridges
- 1 May 2016
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 550-559
Abstract
Reaching Exascale will require leveraging massive parallelism while potentially leveraging asynchronous communication to help achieve scalability at such large levels of concurrency. MPI is a good candidate for providing the mechanisms to support communication at such large scales. Two existing MPI mechanisms are particularly relevant to Exascale: multi-threading, to support massive concurrency, and Remote Memory Access (RMA), to support asynchronous communication. Unfortunately, multi-threaded MPI RMA code has not been extensively studied. Part of the reason for this is that no public benchmarks or proxy applications exist to assess its performance. The contributions of this paper are the design and demonstration of the first available proxy applications and micro-benchmark suite for multi-threaded RMA in MPI, a study of multi-threaded RMA performance of different MPI implementations, and an evaluation of how these benchmarks can be used to test development for both performance and correctness.Keywords
This publication has 8 references indexed in Scilit:
- NiMC: Characterizing and Eliminating Network-Induced Memory Contention2016
- An evaluation of MPI message rate on hybrid-core processorsThe International Journal of High Performance Computing Applications, 2014
- Enabling Efficient Multithreaded MPI Communication through a Library-Based Implementation of MPI Endpoints2014
- Enabling communication concurrency through flexible MPI endpointsThe International Journal of High Performance Computing Applications, 2014
- An Evaluation of One-Sided and Two-Sided Communication Paradigms on Relaxed-Ordering Interconnect2014
- Implementing OpenSHMEM Using MPI-3 One-Sided Communication2014
- Thread-Safe SHMEM ExtensionsLecture Notes in Computer Science, 2014
- The impact of hybrid-core processors on MPI message rate2013