Brief Announcement: (Near) Zero-Overhead C++ Bindings for MPI
Open Access
- 17 June 2024
- conference paper
- Published by Association for Computing Machinery (ACM)
- p. 289-291
Abstract
The Message-Passing Interface (MPI) and C++ form the backbone of high-performance computing and algorithmic research in the field of distributed-memory computing, but MPI only provides C and Fortran bindings. This provides good language interoperability, but higher-level programming languages make development quicker and less error-prone. We propose novel C++ language bindings designed to cover the whole range of abstraction levels from low-level MPI calls to convenient STL-style bindings, where most parameters are inferred from a small subset of the full parameter set. This allows for both rapid prototyping and fine-tuning of distributed code with predictable runtime behavior and memory management. Using template-metaprogramming, only code paths required for computing missing parameters are generated at compile time, which results in (near) zero-overhead bindings.Keywords
Funding Information
- European Research Council (882500)
- Ministerium für Wissenschaft, Forschung und Kunst Baden-Württemberg (33-7533.-9-10/20/2)
- Gauss Centre for Supercomputing
This publication has 5 references indexed in Scilit:
- KaMPIng: Flexible and (Near) Zero-Overhead C++ Bindings for MPI2024
- Towards Modern C++ Language Support for MPI2021
- A large-scale study of MPI usage in open-source HPC applications2019
- Sequential and Parallel Algorithms and Data Structures2019
- Thrill: High-performance algorithmic distributed batch data processing with C++2016