Super Computing and Distributed Systems Camp
Catay, Santander, Colombia
August 15-22, 2010
News & LecturesLecture, 15-07-2010 (14:00 COT)
Observation of Embedded Software Platforms
by Carlos Hernan Prada Rojas, PhD. student University of Grenoble
live stream here
The design of multi-processor systems-on-chip (MPSoC) is a notoriously complex issue. Manufacturers have to integrate new hardware technologies, to develop new system software and to provide new sophisticated functions in a very short time to market. In order to produce optimal software, developers needs of tools for debugging and observation. Tools have also to abstract their features to deal with different and changing hardware and even software platforms. We will focus on techniques for observing embedded systems and we will depict an observation approach based on software components. Such an approach is a possible solution to deal with platform heterogeneity, providing genericity, and partial observation on large and complex systems.
Electromagnetic Modeling in Grid Computing Environments
by Fadi Khalil, PhD. CNRS, LAAS - University of Toulouse
live stream here
Electromagnetic simulation has became a powerful tool in modeling high-frequency and high-speed circuits and devices on digital computer to obtain numerical results. Neverthless, the research is focused on the miniaturization of electronic components and enhancement of performances. This could lead to multiple design parameters to handle, consequently the number of simulations required for the analysis of all possible different configurations may become prohibitive. Besides, these simulations are often complex and time-consuming. Grid Computing, as a solution to these problems, were investigated in a research project entitled MEG, for "Multiscale modeling: from Electromagnetism to Grid". In this talk, feedback on that experience will be presented and discussed.
Accepted students were notified check your e-mail
Self Adaptive Very Large Distributed System
by Filip Krikava from I3S_Modalis Team at Sophia Antipolis, France
live stream here
Despite significant efforts to achieve reliable Grid middlewares, grid infrastructures still encounter important difficulties to implement the promise of ubiquitous, seamless and transparent computing. Acknowledging the fact that middleware can hardly achieve complete reliability in its challenging context, new operation modes have to be implemented to make grid systems resilient and capable of recovering from unexpected failures. In this talk I would like to present an alternative paradigms and techniques that are based on principles used by biological system or in control engineering generally referred to as Autonomic Computing and discuss how they can help to make Grids
Filip Krikava is a PhD candiate at join laboratory of University of Nice Sophia-Antipolis and French National Centre for Scientific Research. He has received his master degree from the Czech Technical University. Prior to starting his research he worked for several years as a software engineer and a technical lead at several companies mainly focusing on developing data intensive distributed applications. His research interests are in large distributed systems, autonomic computing, grid computing, model-driven engineering and service-oriented software architectures.