Parallel Sessions - Approaches to Modeling 4             Event Type Start Time End Time Rm # Chair     Education 3:30PM 5:00PM 21, 22, 23, 24, 25, 28   Title: Approaches to Modeling   Speakers/Presenter:   Education 3:30PM 5:00PM 21, 22, 23, 24, 25, 28   Title: Cluster Computing: Intermediate Message Passing Interface   Speakers/Presenter: Paul Gray, David Joiner   Education 3:30PM 5:00PM 21, 22, 23, 24, 25, 28   Title: GIS2K (GIS to Knowledge)   Speakers/Presenter: Michael Welge, Loretta Auvil, Peter Bajscy   Education 3:30PM 5:00PM 21, 22, 23, 24, 25, 28   Title: Introduction to Parallel and Distributed Computing   Speakers/Presenter: Alf Wachsmann

Session: Parallel Sessions - Approaches to Modeling 4   Title: Approaches to Modeling   Chair:   Time: Sunday, November 16, 3:30PM - 5:00PM   Rm #: 21, 22, 23, 24, 25, 28   Speaker(s)/Author(s):           Description:   Civic Center Rooms [to be assigned] * Agent Modeling – Michael Novak, Lisa Bievenue, Edna Gentry (repeat of 8:30 AM session) * Systems Dynamics Modeling – Shawn Sendlinger, Susan Ragan (repeat of 8:30 AM session) * Algebraic Modeling – Garrett Love, Dan Warner (repeat of 8:30 AM session) * Geowall Demonstration * GIS2K (GIS to Knowledge) – Michael Welge, Loretta Auvil, Peter Bajscy * Cluster Computing: Intermediate Message Passing Interface – Paul Gray, David Joiner * Introduction to Parallel and Distributed Computing - Alf Wachsmann   Link: --

Session: Parallel Sessions - Approaches to Modeling 4   Title: Cluster Computing: Intermediate Message Passing Interface   Chair:   Time: Sunday, November 16, 3:30PM - 5:00PM   Rm #: 21, 22, 23, 24, 25, 28   Speaker(s)/Author(s):     Paul Gray, David Joiner       Description:   This session will apply basic MPI skills to the parallelization of a numerical model. Participants will be introduced to MPI features for collective communication and reduction, as well as debugging tools available for the most common MPI implementations.   Link: --

Session: Parallel Sessions - Approaches to Modeling 4   Title: GIS2K (GIS to Knowledge)   Chair:   Time: Sunday, November 16, 3:30PM - 5:00PM   Rm #: 21, 22, 23, 24, 25, 28   Speaker(s)/Author(s):     Michael Welge, Loretta Auvil, Peter Bajscy       Description:   Geographic Information System (GIS) is a composite of hardware and software used for storage, retrieval, mapping, and analysis of geographic data. Spatial features are stored in a coordinate system (latitude/longitude, UTM, etc.), which references a particular place on the earth. Spatial features and their associated descriptive attributes are combined in the same coordinate system and layered together for mapping and analysis. Applications of GIS come from many scientific domains including land use studies, hydrology, geology, soil management, environmental or development planning. This session will describe analyses of GIS data for territorial partitioning. We will present a suite of tools that use modules in D2K for processing raster and vector data, georeferencing heterogeneous data, extracting statistical features per boundary, aggregating boundaries based on selected features and performing geographical error analysis of territorial aggregations. We will demonstrate our analyses with the vector data defined by the U.S. Census Bureau, USGS terrain elevation maps, forest data, and crime data. This session will overview some of the GIS challenges and illustrate how GIS analysis can be accomplished for territorial partitioning purposes.   Link: --

Session: Parallel Sessions - Approaches to Modeling 4   Title: Introduction to Parallel and Distributed Computing   Chair:   Time: Sunday, November 16, 3:30PM - 5:00PM   Rm #: 21, 22, 23, 24, 25, 28   Speaker(s)/Author(s):     Alf Wachsmann       Description:   “Introduction to Parallel and Distributed Computing" tries to connect general computing with the SuperComputing conference by explaining why parallel computing is a good idea. We will explain how parallel computing can speed up computations. We will show how to parallelize the addition of numbers. After introducing the concept of computation and communication costs and explaining the Big-O notation we will talk about "latency" and "bandwidth." These concepts are then used to explain why writing efficient parallel programs is so hard. We will explain the difference between parallel and distributed computing by introducing the notion of tightly vs. loosely coupled systems, fine vs. coarse grain parallel programs, and synchronous vs. asynchronous computation. We will explain which classes of algorithms (matrix and the like) are well suited for parallel and distributed computing.   Link: --