Augias: A Model-Based Software Development Kit for High-Performance Collaborative and Multiphysics Scientific Computing

The design of simulation studies for HPC is a complex and time-consuming activity, especially when developing multiphysics designs of experiments. Using integrated environments limits the simulation study management efficiency, interoperability, and traceability in a multi-user context. This paper p...

Full description

Saved in:
Bibliographic Details
Published in2023 Congress in Computer Science, Computer Engineering, & Applied Computing (CSCE) pp. 2548 - 2555
Main Authors Nassiet, Didier, Gaisset, Valentin, Lugato, David
Format Conference Proceeding
LanguageEnglish
Published IEEE 24.07.2023
Subjects
Collaboration
Computational modeling
Computer architecture
Data models
Domain-Specific Modeling Language
High-Performance Computing
Interoperability
Model-Driven Engineering
OSGi
Scientific computing
Simulation Lifecycle Management
Software
Online AccessGet full text

Cover

More Information
Summary:The design of simulation studies for HPC is a complex and time-consuming activity, especially when developing multiphysics designs of experiments. Using integrated environments limits the simulation study management efficiency, interoperability, and traceability in a multi-user context. This paper presents the Augias toolkit, a collaborative approach to managing HPC data and processes which addresses these issues by providing complete and neutral modeling of the simulation domain and a service layer end users can extend directly. The Augias metamodels and services are detailed in addition to the technology stack required to implement its principles and architecture. In conclusion, we present the results of an empirical evaluation study confronting the Augias SDK with three other solutions and scoring them through a canonical use case.
DOI:10.1109/CSCE60160.2023.00409