Sites Inria

CEA-EDF-Inria - Computer Science Summer school

Building the next generation of visualization software

From June 13 to June 24  2011, CEA-EDF-Inria, Computer Science Summer school  will organize a course on  "Building the next generation of visualization software".

  • Date : 13/06/2011 au 24/06/2011
  • Lieu : CEA Cadarache, 13108, Saint Paul-lez-Durance, France
  • Intervenant(s) : Hank Childs, Kwan-Liu Ma, Ivan Viola
  • Organisateur(s) : Laurent Colombet, CEA, Thierry Carrard, CEA

Lectures organization

This summer school is organized  following 3 main lecturers:
1 Hank Childs (Lawrence Berkley National Lab)
2 Kwan-Liu Ma (UC Davis)
3 Ivan Viola (Univ. Bergen)

Course 1: How to build a visualization application for very large data.
Speaker:  Hank Childs (Lawrence Berkley National Lab)
        
This course will focus mainly on visualization software architecture and development. It has been decided that topics such as in-situ, I/O issues and remote visualization are part of this course. Those topics could be also covered in the course 2, but in this course the goal is to treat those items from the point of view of what issues they imply regarding the software architecture and development.
Life cycle and sustainability of software tools: which techniques to improve or guarantee development robustness. How to decrease dependencies w.r.t low-level software evolutions (multi-thread or GP-GPU capability, connection with new open source modules, …)
Design and performances: making software framework suitable for efficient use of new HPC architectures.
Algorithms and methods: processor-oblivious and cache-oblivious techniques, many-cores, stream processors.

Course 2: Advanced techniques for visualizing large, complex data
Speaker: Kwan-Liu Ma (UC Davis)
This course will focus mainly on: what will be the visualization concept to deal with petascale simulations.

  • Strategies: How to overcome the ever increasing database sizes through different approaches: multi-resolution, in situ, streaming and adaptive processing.
  • Hybrid computing: Making use of different types of computing units in the same tool. Adaptive frameworks to efficiently draw benefit from available hardware.
  • How to take into account fault tolerance?

Course 3: User-Centric scientific visualization
Speaker :   Ivan Viola (Univ. Bergen)

  • New rendering techniques for better physics insight
  • Perceptual cues
  • Developers’ and users’ feedback about existing tools and their functionalities
  • Data mining, machine learning and features extraction

Mots-clés : CEA EDF Inria Computer Science Summer school Visualization software Petascale

Haut de page