CISTER-lead European project P-SOCRATES meets in Porto
17, Nov, 2014
On September 30th/October 1st, the partners of the European Project P-SOCRATES gathered at CISTER, for two days of intensive technical work. This included in-depth discussions of the scientific and technical work being carried in the project work packages, but important, the specification of the holistic system model and software stack, which will be presented for the 2nd milestone of the project, in the first quarter of 2015. The partners also evaluated the detailed feedback received by the commission on the contents of the first review, which provided very positive remarks on the work being developed, confirming the relevance of the foreseen objectives, as well as stressing the capability and importance of innovation and exploitation.
The P-SOCRATES project is researching and developing new techniques for exploiting the massively parallel computation capabilities of next-generation many-core embedded platforms in a predictable way. These platforms are well positioned for intercepting the increasingly convergence of High-Performance Computing (HPC) and Embedded Computing (EC) domains need for predictable high-performance, allowing HPC and EC applications to be executed on efficient and powerful heterogeneous architectures integrating general-purpose processors with many-core computing fabrics.
The project partners include as research institutions, besides CISTER/INESC-TEC, the Barcelona Supercomputing Centre (Spain), the University of Modena (Italy) and the Swiss Federal Institute of Technology Zurich (Switzerland). The industrial partners of the project include ATOS (Spain) and the SMEs Evidence (Italy) and Active Technologies (Italy). The project partners are supported by an industrial advisory board, which includes well-known multi-national companies including Airbus, IBM, and Honeywell.
Besides overall coordination and technical management, CISTER/INESC-TEC is also deeply involved in the parallelism to real-time activity, leading in particular the Timing and Schedulability analysis work package.