Menu:

Latest news:

30.03.2010:
Stéfano got his Master!

15.03.2010:
ERAD 2010 in Passo Fundo.

19.01.2010:
Mission in Berlin.

20.12.2009:
João Lima has been accepted in PhD.

Links:

Instituições

Research Projects in HPC and Parallel Computing

International Projects

  1. GBRAMS-AMSUD (2008-2009) – CAPES/INRIA (STIC-AMSUD). Coord. Philippe O. A. Navaux - This international project joins together Brazil (UFRGS), Argentina (UBA), Peru (Univ.of Trujillo) and France (project MESCAL/INRIA) to assemble a computational grid of four clusters among our countries, in order to simulate a twenty years climatology for the whole Southern America. The software OAR will be used to run the Grid.
    More information.

  2. CAPA (2008-2010) – CNPq-INRIA. Coord. Philippe O. A. Navaux - This project brings fundings to the common research with the Parallel Processing Group of Grenoble/LIG in High-Performance Computing.
    More information.

  3. BRAFITEC Scistema (2007-2008) – CNPq. Coord. Philippe O. A. Navaux - This project deals with students exchange, at the undergraduate level, with the universities of Grenoble (France). On our side, both our courses of Computer Engineering and Computer Science are involved. Each year, some 5 of our students are sent for two semesters to Grenoble, and we receive around 4 or 5 French students. The SCISTEMA project is the sequel of the former Brafitec project called Paginer (coord.: Prof. Geyer), which has worked on the same basis from 2003 to 2006. It is notable that this kind of interaction with Europe has started as an evolution of the other many research projects that have been going on for more than two decades.
    More information.

  4. Probral (2007-2009) – CAPES-DAAD. Coord. Nicolas Maillard / Ph. O. A. Navaux - Sets up a collaboration with the German partners from the Technische Üniversität Berlin and the GPPD, besides Cesar A. F. De Rose’s group (PUCRS, Brazil). The fields of interest are the management of Grids and Clusters, and parallel programing. Two Brazilian PhD students are currently spending one year of their studies in Berlin, within this project. The GPPD has received the visit of the PhD student Jörg Schneider, during 5 weeks in 2007. It is notable that the scientific collaboration with the group of the Prof. Heiss has also resulted in a UNIBRAL project, for under-graduate students exchange.
    More information.

  5. Équipe Associée DIODE-A (2006-2011) – INRIA. Coord. Bruno Raffin - The research group where I am working (GPPD) has been chosen as “Équipe Associée” (Associated Team) by the French national organism of research in Computer Science INRIA, in Dec., 2006. As a comparison, there are only two associated teams in Brazil, and 27 worldwide.
    More information.

  6. P(AD)2 (2006-2009) – CAPES-COFECUB. Coord. Philippe O. A. Navaux - Sets up a collaboration with the French INRIA projects called MOAIS and MESCAL, with exchange of Master and PhD students. The main themes are Grid Computing, Parallel Programming (MPI, OpenMP, Kaapi) and distributed file-systems. Until now, two Brazilian PhD students have spent one year in France within P(AD)2, and the GPPD has received 2 PhD students from Grenoble.
    More information.

  7. PICS (2005-2007 OVER) – CNPq/CNRS. Coord. Bruno Raffin - This project has fostered the common research between the GPPD and the parallel computing group of Grenoble/LIG in High-Performance Computing (IO systems, Monitoring, Adaptive programming with MPI). The GPPD has received a counterpart budget from the CNPq, in 2007 (coord.: Prof. Philippe O. A. Navaux).
    More information.

National Projects

  1. Projeto Multicore (2008-2009) – CNPq (Universal). Coord. Philippe O. A. Navaux - In partnership with the Profs. Navaux and Luigi Carro, this project consists in the study of the new parallel architectures and of how they can be programed. The study has started with the support that reconfigurable architectures can provide to run efficiently PRAM programs. This project includes a PhD collaborator, two PhD students and three Master students.
    More information.

  2. RECLIRS (2007-2009) – FINEP (Redes Estatuais). Coord. Haroldo Fraga dos Campos Velho - This project is a regional offspring of the GBRAMS project, local to the state Rio Grande do Sul. Together with the colleagues from meteorology and hydrology (IPH/UFRGS), and also with groups from Pelotas, from the INPE/Santa Maria and from the Federal University of Santa Maria, we are simulating the evolution of the climate in our state.
    More information.

  3. Interoperability (2007-) – Microsoft. Coord. Philippe O. A. Navaux - This project, in partnership with Roberto Prado (Microsoft), includes quite a few research lines under the supervision of four researchers from the UFRGS. I am responsible for the sub-project called MPI#, which deals with the integration of MPI and C#. I am also supervising the work on a tool that compiles ODT documents into OpenXML ones. These two lines of work include two undergraduate students (one of them making his concluding work), and a Master student.
    More information.

  4. Massive Atmosphere (2007-2008) – CNPq (Grandes Desafios). Coord. Philippe O. A. Navaux - Together with the CPTEC (Dr. Jairo Paneta) and the LNCC (Prof. Pedro Dias), we are working in this project on the study of the impact of Multicore architectures in the performance of the parallel programs (using MPI and/or OpenMP) that are used to model and forecast the weather and the climate in Brazil. These software include BRAMS and OLAM.
    More information.

  5. Curriculum for the Intel Multi-core technology (2006-2007 OVER) – INTEL. Coord. Philippe O. A. Navaux - Has dealt with the programming of Multicore architectures. The aim was to foster the teaching of the technology and its programming tools (OpenMP and MPI). We have provided lectures, and the setup of a laboratory of 10 dualcores PCs, besides some talks in national events. The project has also involved the Prof. Carissimi and two PhD students.
    More information.

  6. G-BRAMS (2005-2007 OVER) – FINEP. Coord. Philippe O. A. Navaux - This project has dealt with the use of a Computational Grid, in collaboration with the partners from the INPE (LAC e CPTEC), in order to test such a platform with a production-like software of climatological simulation, called BRAMS. BRAMS is based on MPI. Three middleware for Grid computing have been compared: Globus, OurGrid and OAR. This project has also had the support of HP Brazil.
    More information.

  7. JAVA-WSPAD (2005-2007 OVER) – FINEP. Coord. Philippe O. A. Navaux - In partnership with Prof. Claudio Amorim, from the COPPE (Univ. Federal do Rio de Janeiro). This project has focused on the use of Web Services to set up a cluster of clusters and unify the management tools (job submission, batch scheduler, monitoring, etc.). Two other researchers of the GPPD have worked on this project, besides 3 graduate students. The project has counted with the support oh HP Brazil.
    More information.

  8. CLUMSSY (2004-2006 OVER) – HP Brasil. Coord. Philippe O. A. Navaux - In this project, I have led the research on scheduling of parallel, MPI programs. We have done it in two contexts: statically (off-line), by using the communication graph of a former execution to map the processes to the processors; and dynamically, by spawning new processes at run-time to balance the load.
    More information.


    —
    Fri Apr 02 16:48:12 -0300 2010