Grupo de Ingeniería Microelectrónica

Grupo de Ingeniería Microelectrónica

Departamento de Tecnología Electrónica, Ingeniería de Sistemas y Automática Universidad de Cantabria
Home   Personas   Investigación   Docencia   Doctorado   Publicaciones   Herramientas   Bolsa de Empleo   english version Sat 20-Apr-24 . 13:00

Mapa Web



Info Santander

Gestión BD

   Ficha completa
Título:Automatic Concurrency generation through Communication Data Splitting based on UML-MARTE Models
Tipo:Publicacion en Proceedings o Actas internacionales
Lugar:XXVIII Conference on Design of Circuits and Integrated Systems, San Sebastian, Noviembre, 2013
Autores: Alejandro Nicolás
Héctor Posadas
Pablo Peñil
Eugenio Villar
Líneas: Diseño y verificación de sistemas embebidos HW/SW
Proyectos: FP7 288307 PHARAON
Fichero:ver fichero
Resumen:With the increase in the number of processing elements integrated in HW platforms, the development of new solutions helping engineers to design concurrent applications is gaining greater interest. Code refinements required to parallelize sequential algorithms are usually
quite complex, and do not guarantee that the decisions taken will provide the expected performance result, requiring a costly and iterative design process. To solve this problem, this paper presents a methodology that enables designers to define and automatically create different concurrent architectures for the system, only modifying different parameters in the UML model. The main idea is to
automatically modify the communication between components, adding to the typical configurable options of performing synchronous, asynchronous or buffered calls the idea of enabling parallelization by dividing one service call into several concurrent calls, each one operating with part of the data. Examples of use of this idea with video applications are presented, demonstrating the performance
improvement obtained in the final system.
© Copyright GIM (TEISA-UC)    ¤    Todos los derechos Reservados.    ¤    Términos LegalesE-Mail Webmaster