Heterogeneous cluster environments are becoming an in-creasingly popular platform for executing parallel applica-tions. Efficient heterogeneous parallel applications must account for the differences inherent in such an environ-ment. Specifically, faster machines should possess more data items than their slower counterparts and communi-cation should be minimized over slow network links. We propose the -Heterogeneous Bulk Synchronous Parallel (HBSP ) model, which is based on the BSP model of com-putation, as a framework for developing applications for heterogeneous systems. The BSP model is appropriate for-level (one communication network) heterogeneous sys-tems. HBSP extends BSP hierarchically to address -level heterogeneous machines. The utility of the model is demon-strated through the design and analysis of the gather and one-to-all broadcast operations. Our results indicate that the HBSP model guides the efficient design of heteroge-neous parallel applications in an architecture-independent manner. 1.

In this paper, we present a recursive molecular model suitable for multiresolution analysis in a variety of application areas. Our approach allows ease of use by computer scientists and biologists by proposing a software interface for molecular analysis that hides programming details. We demonstrate that our design is flexible enough for use in desktop analytical applications, in large-scale parallel simulations, and as a server for remote molecular analysis across wide-area networks.