RESOURCE-INDEPENDENT EXECUTION SUPPORT ON LARGE SCALE SYSTEMS
With the movement to multi-core technologies, processor evolution is making rapid progress: more and more cores are being integrated into single processors, more and more machines get connected with increasing bandwidth, and processors become heterogeneous and reconfigurable, allowing for dynamic adaptation to specialized needs. This hardware development is currently moving faster than the software development, rendering application execution inefficient due to the lack of an appropriate Operating System architecture.
The S(o)OS project addresses future distributed systems on the level of a holistic operating system architecture by drawing from Service Oriented Architectures and the strength of Grids. This will decouple the OS functionality from the underlying resource infrastructure, making adaptation to new infrastructures fast and easy, and allowing efficient program execution across an almost unlimited number of varying devices.
S(o)OS investigates means to execute distributed programs in a way that addresses both code requirements and resource availability, thus improving overall performance. The project has already provided a first OS architecture and is now developing generic methodologies for future Operating Systems. It will also elaborate how the specific functionalities and capabilities affect and can be exploited by developers, as well as future processor manufacturers. First results and experimental tools are available from the website.
S(o)OS is funded by the European Commission under Grant Agreement No. 248465, in the context of the TERACOMP initiative.
The project consortium is formed by international experts in the area of distributed and parallel computing:
- High Performance Computing Centre (HLRS), University of Stuttgart, Germany
- Ecole Polytechnique Fédérale de Lausanne, Switzerland
- Europäisches Microsoft Innovations Center GmbH, Germany
- Instituto de Telecomunicacoes-Pólo Aveiro, Portugal
- Scuola Superiore di Studi Universitari e di Perfezionamento Sant‘ Anna, Italy
- Computer Architecture for Embedded Systems (CAES), Universiteit Twente, The Netherlands
Höchstleistungsrechenzentrum Universität Stuttgart, Nobelstraße 19, 70569 Stuttgart, Germany
Phone: ++49-711-685-87262, Fax: ++49-711-685-77262