The ARGO project, being coordinated by Karlsruhe Institute of Technology (KIT), is working on a method to convert model based applications into multicore optimised C code with guaranteed real-time constraints.
Today, says the project, programmers have to adapt their code to the target hardware architecture, which brings high cost and prevents the code from being transferred to other architectures.
“Two of the most important requirements of future applications are increased performance in real time and further reduction of costs without adversely affecting functional safety,” said project coordinator Professor Jürgen Becker from KIT. “For this, multicore processors have to make available the required performance spectrum at minimum energy consumption in an automated and efficiently programmed manner.”
Programming of heterogeneous multicore processors is complex and these programs have to be tailored precisely to the target hardware.
“Under ARGO, a new standardisable tool chain for programmers is being developed,” said Prof Becker. “Even without precise knowledge of the complex parallel processor hardware, programmers can control the process of automatic parallelisation in accordance with the requirements. This results in a significant improvement of performance and a reduction of costs.”
In the future, says the team, the ARGO tool chain could be used to manage parallelisation complexity and adaptation to the target hardware in a largely automated manner and at relatively low cost.
Other partners in ARGO are: University of Rennes (France); Technological Educational Institute of Western Greece; German Aerospace Centre; Fraunhofer Institute for Integrated Circuits IIS; Recore Systems; Scilab Enterprises; and AbsInt Angewandte Informatik.