Towards universal MPI bindings for enhanced new language support

In the field of High Performance Computing (HPC), Message Passing Interface (MPI) is the most widely used and prevalent programming model. Only the low-level programming languages C, C++, and Fortran have bindings available in the standard. Although there are attempts to provide MPI bindings for other programming languages, these may be limited, which could lead to incompatibilities, performance overhead, and functional gaps. To address those problems, we present MPI4All, a brand-new tool designed to make the process of developing effective MPI bindings for any programming language more straightforward. Support for additional languages can be added with little difficulty, and MPI4All is independent of the MPI implementation. Programming language binding generators for Go and Java are included in the most recent version of MPI4All. We demonstrate their good performance results with respect to other state-of-the-art approaches. This work is an extended version of the ICCS-2024 conference paper (Piñeiro et al., 2024). • MPI4All simplifies the development of MPI bindings for any programming language. • Ensures full MPI API coverage, including latest standards, with minimal input. • Automatic updates ensure compatibility with future MPI versions in seconds. • Go and Java show competitive performance compared to state-of-the-art solutions.

Authors

Institutions

Publication Details

Journal
Journal of Computational Science
Published
2025-03-13
DOI
https://doi.org/10.1016/j.jocs.2025.102557
Citations
1
Primary Topic
Parallel Computing and Optimization Techniques
Type
article
Field-Weighted Citation Impact
2.27

Funders

Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
article

Towards universal MPI bindings for enhanced new language support

Juan C. Pichel, Álvaro Vázquez, César Piñeiro
1 citations
Journal of Computational Science
Parallel Computing and Optimization Techniques
2.27
article

Towards universal MPI bindings for enhanced new language support

Juan C. Pichel, Álvaro Vázquez, César Piñeiro
article en
1 citations

Abstract

In the field of High Performance Computing (HPC), Message Passing Interface (MPI) is the most widely used and prevalent programming model. Only the low-level programming languages C, C++, and Fortran have bindings available in the standard. Although there are attempts to provide MPI bindings for other programming languages, these may be limited, which could lead to incompatibilities, performance overhead, and functional gaps. To address those problems, we present MPI4All, a brand-new tool designed to make the process of developing effective MPI bindings for any programming language more straightforward. Support for additional languages can be added with little difficulty, and MPI4All is independent of the MPI implementation. Programming language binding generators for Go and Java are included in the most recent version of MPI4All. We demonstrate their good performance results with respect to other state-of-the-art approaches. This work is an extended version of the ICCS-2024 conference paper (Piñeiro et al., 2024). • MPI4All simplifies the development of MPI bindings for any programming language. • Ensures full MPI API coverage, including latest standards, with minimal input. • Automatic updates ensure compatibility with future MPI versions in seconds. • Go and Java show competitive performance compared to state-of-the-art solutions.

Journal of Computational ScienceVol. 87
Universidade de Santiago de Compostela (ES)
Ministerio de Ciencia e Innovación, Xunta de Galicia, European Regional Development Fund
Quality Education
Openalex Percentile: Top 17%
Parallel Computing and Optimization Techniques
2.27
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.