Issue No. 04 - July (1996 vol. 16)
DOI Bookmark: http://doi.ieeecomputersociety.org/10.1109/38.511859
Metacomputing entails the combination of diverse, heterogeneous elements to provide a seamless, integrated computing service. We describe one such metacomputing application that uses Compositional C++ to integrate specialized resources, high-speed networks, parallel computers, and VR display technology to process satellite imagery in near-real-time. From the virtual environment, the user can query an InputHandler object for the latest available satellite data, select a satellite pass for processing by CloudDetector objects on a parallel supercomputer, and have the results rendered by a VisualizationManager object that could be a simple workstation, an ImmersaDesk, or a CAVE. With an ImmersaDesk or CAVE, the user can navigate over a terrain with elevation and through the cloudscape data as it is being pumped in from the supercomputer. We discuss further issues for the development of metacomputing capabilities with regards to the integration of run-time systems, operating systems, high-speed communication, and display technologies.
Metacomputing, parallel and distributed languages, high-speed networks, virtual environments.
Carl Kesselman, Craig A. Lee, Stephen Schwab, "Near-Real-Time Satellite Image Processing: Metacomputing in CC++", IEEE Computer Graphics and Applications, vol. 16, no. , pp. 79-84, July 1996, doi:10.1109/38.511859