Remote Visualization of Large Datasets

CoReViz Data sheet (PDF)

Modern scientific simulations generate large datasets at remote sites with appropriate resources (supercomputers and clusters). Bringing these large datasets to the computers of all members of a distributed team of collaborators is often impractical or even impossible: there might not be enough bandwidth, storage capacity or appropriate data analysis and visualization tools available locally.

The design of CoReViz is based on the experience of researchers at Tech-X Corporation of working with distributed collaborations. As an example, a researcher might run some simulation or modeling application on a high-performance computer at a remote location. The generated data now resides at a remote site. The researchers use multiple tools such as IDL from Research Systems to extract particular subsets of the data and create 2D or 3D objects, which they then transfer for local viewing.Cost Effective CoReViz allows users to view remote simulation data that may be too large to be moved across the Internet efficiently and/or require applications or resources only available on supercomputers.

diagram of CoReViz high level architecture

CoReViz partially automates such workflows and provides an integrated environment for data visualization, allowing multiple visualization tools. CoReViz allows users to leave the raw data at the location where it was generated, automatically collocate data analysis and visualization routines coming from multiple locations, run these routines on the data, and automatically distribute visualization results to multiple collaborators. The visualization objects and the data analysis and visualization scripts are much smaller than the raw data, thus the network is not overburdened.

Multi-user collaboration environment

CoReViz allows users to view remote simulation data that may be too large to be moved across the Internet efficiently and/or require applications or resources only available on supercomputers. Additionally, CoReViz allows multiple users to collaborate on that data. In multi-user mode, the user is presented with the main client window, after authentication and connection to the CoReViz server. This window lists all of the current users who are connected to the server. This server provides simple ways to communicate with the users with open sessions: one can broadcast messages or send messages to a particular person using a simple chat facility.

Share annotations via markup capability

The image-handling window allows a user to distribute images, or share simple markups and clicks. These features are helpful in discussion of scientific results when one person wants to highlight a particular detail of the image.

Manipulation of 3D objects

CoReViz gives researchers the ability to analyze data and create visualizations at remote sites by offering plug-in capability for visual tools like IDL and OpenDX, and using OpenGL for representation and VRML for viewing. When displaying VRML files, users may choose to share the view as they rotate, pan and zoom in real-time, or only when they decide to send an update to the other users in the session.

Movie generation and sharing

Animation has become increasingly useful for studying dynamic phenomena. Researchers routinely create movies to capture a dynamic system at different time steps. To help automate this process and to be able to share the animation, CoReViz has movie generation capabilities which produces QuickTime movies. Once a movie is created and distributed, all viewers can synchronously view it and use actions such as pause and restart to control the process.

Supports multiple platforms and standard hardware

The CoReViz client, written in Java, will execute on Windows 2000/XP, Linux with kernel 2.x, or Macintosh OS X. The CoReViz server, written in C++ and Python, has been ported to run on Linux and AIX. The data objects rendered for collaborative visualization use standard outputs such as JPEG, PNG and GIF files for 2D visualization objects, VRML files for 3D visualization objects, and QuickTime movies for animation objects.