Skip to main content

Computing large objects visualisation information system


The goal of the project is to develop a portable and efficient tool for the post-processing and real-time visualisation of very large datasets including those too large to output to a network or to store in a raw state in data repositories.

Previously it has not been possible to visualise datasets with more than 256{3} entities within an acceptable response time. Using data compression, decompression and reduction techniques, The project will make it possible to process and visualise very large datasets within a framework open to existing and emerging standards such as OpenGL, PVM and MPI. The target dataset size to be processed is about 1000{3} entities taking into account hardware capabilities, memory, I/O bandwidth, CPU and graphics speeds. The project will address currently available multiple-processor architectures including MPP systems, workstation clusters, vector processors, super-servers and high-end workstations, with optimised implementations using standard high performance tools such as MPI, PVM and OpenGL.

Using the results of the project, engineers will evaluate their work, exploiting efficiently the results of simulations in a number of areas such as computational fluid dynamics, structural analysis, reservoir simulation, molecular simulation and large-scale discrete and continuous data analysis. The project will address the advanced research, industrial simulation, and educational markets.


Tethys Sa
Immeuble Le Capricorne, Rue Colbert 23
78885 St Quentin En Yvelines

Participants (2)

Intecs Sistemi S.P.A.
Via L. Gereschi 32/34
56127 Pisa
United Kingdom
Vineyard Chambers 3
OX14 3PX Abingdon