• ModViz's http://www.modviz.com/products Virtual Graphics Platform, an OpenGL cluster visualization system that dynamically instantiates render, composite, and display components as required;
• Sun Microsystems' http://www.sun.com/products-n-solutions/ hardware/docs/ Servers/compute_grid/sun_fire_visual_grid/index.html Sun Fire Visual Grid product, which networks high-end Sun graphics hardware; and
• IBM's http://www-03.ibm.com/servers/deepcomputing/visualization Deep Computing Visualization hardware, which aims at both cluster rendering and remote visualization.
• assign resources for a tiled screen environment in which a single cluster node feeds each screen and
• enhance the resolution or frame rate on a single machine.
• The conceptual layer loosely explains the structure for visualization and collaboration.
• The logical layer maps the concepts to software entities.
• The physical layer maps the logical structure to a Grid environment.
• The Chromium configuration data deals with how to configure each server as part of the pipeline.
• The additional server data deals with how the system will represent the physical environment (such as position in a tiled display).
• The server status data includes scheduling information relating to external factors such as machine load.
• the client node, on which the jgViz client software runs;
• the application node, on which the Chromium application faker hosts the application;
• rendering slave nodes, which render OpenGL commands and then either display the results locally (tiled pipelines) or read back the results and send them over the network to a compositor node (readback pipelines); and
• the compositor node, which receives multiple rendering streams, reassembles them, and displays them locally (readback pipeline only).
• stop the old pipeline (involving stopping the mothership and killing the application via Grid jobs),
• refresh metrics data on available nodes,
• transfer the new configuration to the new mothership, and
• launch both the new mothership and application nodes.
Ade J. Fewings is a visualization engineer at Technium CAST (Centre for Advanced Software Technology), a technology-led business incubation environment. His research interests are in visualization in distributed systems, particularly involving Grid and parallel graphics technologies. He received his PhD in computer science from the University of Wales, Bangor. Contact him at Technium CAST, Parc Menai, Bangor LL57 4HJ, Wales, UK; email@example.com.
Nigel W. John holds a chair in computing at the University of Wales, Bangor. His research interests include high-performance visualization and medical graphics. He received his PhD in mathematical sciences from the University of Bath. He's a member of the Eurographics Association and is chair of the Eurographics Professional Board. Contact him at the School of Computer Science, Univ. of Wales, Bangor, Dean St., Bangor LL57 1UT, Wales, UK; firstname.lastname@example.org.