VELaSSCo: Visualization for Extremely Large-scale Scientific Computing

VELaSSCo

Introduction


One of the major problems experienced by engineers and scientists running high-end real-scale simulation models is the management and exploitation (in terms of visualization) of the outputs of their models. Inspecting large dataset distributed over numerous clusters located remotely, is often a significant difficulty to be overcome to ensure that users (both in academia and industry) are able to take the best advantage of current (and future) HPC infrastructures.

The Vision of VELaSSCo is to provide new approaches for visual analysis of large-scale simulations for the Exabyte era. It does this by building on big data tools and architectures for the engineering and scientific community and by adopting new ways of in-situ processing for data analytics and hardware accelerated interactive visualization.

To better manipulate the data from simulations with billions of records it is crucial that the engineering and scientific community adopts big data tools. VELaSSCo will provide a simulation data analysis platform consisting of:

  • A database structure, based on widely used technologies such as Hadoop-HBase, that can organise and store a diverse range of large-scale simulation data sets for collaborative use.
  • An innovative approach, adopting big data best practices, to handle large scale simulation data sets that have to be stored on multiple servers.
  • A framework equipped with advanced in-situ processing tools to analyse the output of parallel simulation solvers.
  • An analysis platform to analyse and visualize large-scale data sets interactively. This builds on leading edge graphics hardware.

Project Partners

The VELaSSCo consortium has a wide variety of expertise aligned with the objectives of the project. Expertise of the partners can be classified mainly in two principal categories: Institutions specialized in the development and implementation of visualization software, particularly addressed to facilitate analysis of HPC simulations in engineering and scientific first-class problems (CIMNE, UNEDIN and FRAUNHOFER); and institutions with experience in Data Analytics, Big Data, Big Data Handling and Cloud Computing (SINTEF, INRIA, JOTNE and ATOS).

The partnership also counts with a large group of potential end-users of the outcomes of the projects, ranging from large engineering companies, SME’s on engineering modelling and software houses, to research organisations doing large scale engineering modelling, linked to the User Panel already described.

List of partners:

Project Gallery


John P. Morrissey
John P. Morrissey
Research Scientist in Granular Mechanics

My research interests include particulate mechanics, the Discrete Element Method (DEM) and other numerical simulation tools. I’m also interested in all things data and how to extract meaningful information from it.

Related