Contents
Editorial
7i program
Tribune
Guest contributors
Hot topics
Business cases
Solutions
Quick Poll
At a glance
Events
PDF version
 

Subscribe to
Bull Direct:


 

Archives
May 2007
Guest contributors
FAME2: digital innovation and simulations
Comment from partners

CEA LIST - Christian Fluhr, Director of Research:
“The multimedia information market, both for business and personal users, is undergoing a huge expansion. Media takeovers – such as the purchase at vast expense of FlickR by Yahoo, or YouTube by Google – show that while multimedia search applications were once the preserve of professional users, the general public is now taking a much bigger interest in them with the advent of Web 2.0. This is accompanied by a considerable increase in the volumes of image data needing to be processed. The technologies used at the present time are rather simple: keywords inserted by authors or documentation providers, and pixel-by-pixel image comparison. In the future, the challenge will be to deliver automated and truly semantic descriptions of the multimedia in question even given the likely volume of data being generated via the Web. This evolution, of course, comes with a price tag. The semantic analysis of images, text, words and sounds requires an enormous amount of processing power to handle the huge volumes involved, and so recourse to large capacity multiprocessor machines such as those designed as part of the FAME2 program.”

Dassault Aviation – Alain Samblat, Head of high-performance computing resources: “Aerodynamic digital simulation is at the core of our design processes, notably for our Falcon business aircraft. The combination of fine modeling of flow physics with high-level computing power has enabled us to revolutionize aerodynamic design over the past few years. Our ambitious performance objectives mean we need to combine intensive utilization of the most recent computing software with state-of-the-art resources and testing techniques. The FAME2 program is enabling us to anticipate the development of simulation resources, and is therefore contributing to the development of our design process and the construction of even more powerful aircraft, that are environmentally friendly while remaining highly competitive.”

IFP, Stephane Requena, Head of high-performance computing resources:
“The increase in world energy demand and the progressive scarcity of petrol and gas, as well as environmental considerations and a problematic road ahead when it comes to finding viable alternatives in the transport and petrochemical sectors, are just some of the factors that are going to shape tomorrow’s energy landscape. Thanks to the FAME2 program, we are able to validate two major applications in this field: simulation of a complete car engine, which will enable us in turn to optimize consumption and reduce emissions of pollutants, and surveys on new oil reserves. FAME2 will allow us to increase the precision of this kind of research by a factor of 20 (112 million grid cells: a world first), enabling increased profitability (of between 10 and 15%) for any given oilfield, with the accompanying improved efficiencies in its exploitation. Uncertainty will soon be a thing of the past in these areas of research!”

IBISC Laboratory at the University of Evry – Fariza Tahi, Assistant Professor, on the theme of: “Towards a data warehouse for integrative biology data: from the genome to man”. “Working with our international partners in the context of the European EuroPhysiome initiative – whose ambition is to analyze the human body as a unique and complex system through the creation of a Virtual Physiological Man – our laboratory is responsible (in collaboration with our international partners) for one of the vital organs, the kidney. As part of the FAME2 program, we are developing a warehouse for storing heterogeneous Renal Physiome data under the XediX system (the CEA’s native XML database). This means biologists will be able to model and simulate renal functioning using saved anatomical and functional data as a starting point; these simulations will make it possible to avoid lengthy and costly traditional experimentation programs. Research teams and medical staff will be able to formulate highly complex query routines and run them through this data warehouse to support diagnosis and treatment anywhere in the world.”

The FAME2 project profile

 
Contact  |  Site map  |  Legal  |  Privacy