The Virtual World Data Server (VWDS) client for large scientific datasets

The UCLA Virtual World Data Server (VWDS) is a terabyte data server with a clustered architecture based on commodity PC's. It provides data storage for multiple, simultaneous, high-performance clients running different multimedia and virtual world applications. It serves the data on demand, in realtime, with a pre-negotiated data transfer rate. UCLA is currently running two prototype servers and building a large VWDS for campus use. Here we present dataView, our client for the navigation of virtual worlds created on the fly from large scientific datasets. dataView allows scientists to fly through large, regularly-gridded, time-varying 3D datasets from their desktop computers. With dataView the scientist can change isosurface levels, colormaps, and other geometric features at will. The data is first preprocessed with a small program that calls several functions from dataView's class library. This divides the data into cubes and sub-cubes (which we refer to as tiles and subtiles), samples it at three levels of detail, and writes it to the VWDS. When dataView runs, the selection of data sub-cubes and the computation of the necessary geometry are performed by client components running in parallel on the VWDS's compute nodes. The scientist's computer is used only for user interaction and rendering. This paper describes the data organization, the client components, and how the system components interact.