Modern scientific research is increasingly dependent on simulation and analysis requiring high performance computers, distributed large-scale data and high-speed networks. This project, Strategies for Remote Visualization on a Dynamically Configurable Testbed, will use an integrated, application-oriented design and development approach to build highly-optimized distributed applications. These applications will use network services as first class resources, together with distributed compute, storage, and visualization resources. In particular, the project will investigate new strategies for distributed visualization of large data, prototyping an advanced distributed volume rendering system combining user-controlled network services,high-speed data streaming, high-quality video streaming and parallel GPU-accelerated interactive rendering. Working with distributed national (LSU, NCSA, ORNL, TACC) and international (Masaryk University) collaborators, as well network providers (Internet2, LONI, OmniPop, ESNET) the project will build a real-world testbed for implementation and evaluation of strategies.
Intellectual Merit This project simultaneously addresses issues in distributed computing and computational science, and is fundamentally targeted at the development and enhancement of a national cyberinfrastructure to enable scientific discovery on increasing large and complex computing environments such as provided by the NSF's TeraGrid, XD, Blue Waters and DataNet programs.
Broader Impact The real-world testbed, utilizing Internet2's Dynamically Configurable Network service, will provide a persistent experimental environment that will be used by a range of application groups to prototype and experiment distributed applications. The testbed will be available to over forty students and postdocs working in the cyberinfrastructure component of Louisiana's NSF EPSCOR Research Infrastructure Improvement award.
