Robert Hollebeek - US grants

9413948 Hollebeek This award is to a consortium group from The University of Pennsylvania, the University of Maryland, and the University of Illinois at Chicago. The consortium is proposing to develop a scalable, wide area high performance meta-computer for problems which require numerically intensive analysis of large amounts of geographically distributed data. The experimental facilities will include a high-speed network linking existing parallel computers and high-performance clusters of workstations at the three institutions. In addition each institution in the consortium will acquire or upgrade a parallel machine. Software will be developed to implement the meta-computer concept whereby an individual using a node at any site will have access to all other nodes and distributed data as needed, regardless of the location of these entities. The meta-computer design will be scalable in that additional distributed nodes can be added at any time. High speed access to the Internet will link the geographically-distant components, and technologies such as ATM will provide the local connections. Research and application areas to be addressed using these facilities include particle physics and condensed matter physics, imaging and virtual reality, and digital libraries. ***

9601861 Hollebeek, Robert Smith, Jonathon University of Pennsylvania Academic Research Infrastructure: Development of a Wide Area Scaleable Infrastructure for Data Intensive Computing: Consortium Proposal This Academic Research Infrastructure award supports the acquisition of equipment to enhance the National Scaleable Cluster Project. The equipment will be sited at 3 universities to further the development of a national meta computing center. The equipment consists of large fast disk arrays, high performance computer cluster enhancements, ATM network switches to OC-12 capabilities, network connections, and associated software. The research projects supported by the equipment include: 1. Applications computing that is data intensive including data mining of particle physics data, digital library development of the linguistics data consortium, digital library development of astrophysics, gravitational lenses data maps, galactic large structure simulations, and the development of numerical and statistical digital libraries. 2. Real time computing applications including collecting reduced astronomical data, medical imaging, and telepresence. 3. Parallel compute intensive applications including liquid crystal imaging, phase transition simulations, crystal surface computational studies, and vortex flow studies. 4. High speed network research including network striping and the inclusion of tertiary store in high performance memory hierarchies. 5. Computer science research on scaleable object stores, research on caching, transforming, and replicating data, and software tools for cluster management.