Improving performance and reducing costs, while maintaining safety and programmability, are the main requirements for embedded systems (e.g. in the aerospace and automotive fields). For many applications, customised computing platforms based on heterogeneous multi-core architectures are the only way to ensure the necessary performance at low power consumption. However, parallel programming with time-critical embedded applications suffers from a complex tool chain and programming process.
The ARGO research project addresses this challenge with a holistic approach to programming for heterogeneous multicore architectures, using automatic parallelization of model-based real-time applications. ARGO aims at improving parallelization that takes into account the predictability of worst-case execution time. The programming approach is to combine automatic parallelization with user-guided parallelization. This reduces the need for programming expertise for heterogeneous architectures.
Imen Fassi, a post-doctoral student, and Thomas Lefeuvre, a doctoral student, have recently joined the permanent team.
The ARGO project is funded by the European Horizon 2020 programme.
It brings together 8 partners from industry and academia:
- Karlsruhe Institute of Technology
- Scilab Enterprises
- Recore Systems
- Université de Rennes 1
- Technological Educational Institute of Western Greece
- AbsInt Angewandte Informatik GmbH
- German Aerospace Center (DLR) Institute of Flight Systems
- Fraunhofer Institute for Integrated Circuits
European project ARGO - HORIZON 2020 Programme
Name of the scientific leader: Steven DERRIEN
Laboratory / UMR: IRISA (Institut de Recherche en Informatique et Systèmes Aléatoires) - UMR CNRS 6074
Project duration: From 01/01/16 to 31/12/18 (3 years)