Title | ||
---|---|---|
Analyzing Mpi-3.0 Process-Level Shared Memory: A Case Study With Stencil Computations |
Abstract | ||
---|---|---|
The recently released MPI-3.0 standard introduced a process-level shared-memory interface which enables processes within the same node to have direct load/store access to each others' memory. Such an interface allows applications to declare data structures that are shared by multiple MPI processes on the node. In this paper, we study the capabilities and performance implications of using MPI-3.0 shared memory, in the context of a five-point stencil computation. Our analysis reveals that the use of MPI-3.0 shared memory has several unforeseen performance implications including disrupting certain compiler optimizations and incorrectly using suboptimal page sizes inside the OS. Based on this analysis, we propose several methodologies for working around these issues and improving communication performance by 40-85% compared to the current MPI-1.0 based approach. |
Year | DOI | Venue |
---|---|---|
2015 | 10.1109/CCGrid.2015.131 | 2015 15TH IEEE/ACM INTERNATIONAL SYMPOSIUM ON CLUSTER, CLOUD AND GRID COMPUTING |
Keywords | Field | DocType |
MPI-3.0, process shared memory, intranode communication, stencil, multicore | Uniform memory access,Shared memory,Computer science,Parallel computing,Stencil,Distributed memory,Stencil code,Data diffusion machine,Optimizing compiler,Distributed shared memory | Conference |
ISSN | Citations | PageRank |
2376-4414 | 3 | 0.43 |
References | Authors | |
9 | 6 |
Name | Order | Citations | PageRank |
---|---|---|---|
Xiaomin Zhu | 1 | 3 | 0.43 |
Junchao Zhang | 2 | 133 | 13.02 |
Kazutomo Yoshii | 3 | 249 | 18.53 |
Shigang Li | 4 | 282 | 43.13 |
Yunquan Zhang | 5 | 327 | 43.92 |
Pavan Balaji | 6 | 1475 | 111.48 |