In order to effectively deploy large-scale simulations across distributed resources, it is vital that design models be developed for design and performance prediction. Currently, many large-scale simulations are being deployed across clusters of machines. Clusters fall into two broad categorizations: homogenous and heterogenous. In homogenous clusters, resources are basically identical. For heterogeneous, resources can have great variations in structure and speciality. In this talk, we will focus on homogeneous clusters, and on hierarchical clusters, a type of heterogenous cluster. We will present and/or introduce various general design models. We will show that our models are realistic and have wide-utility in a variety of arenas: as design models, performance models, and complexity models. Moreover, our models have been validated on a number of applications across a number of engineering and computational science applications, including spin system simulations (for example for quantum chromodynamics), protein folding, numerical linear algebra (for example large-scale system solvers), tissue simulation (for example tumor growth), information retrieval, network centric operations, and social network analyses. Bio: Eunice E. Santos has B.S. and M.S. degrees in both Mathematics and Computer Science, and received a PhD in 1995 from the University of California, Berkeley in Computer Science, From 1995-2000, Dr. Santos was a faculty member in the Department of Electrical Engineering and Computer Science at Lehigh University. Since 2000, Dr. Santos is the Director of the Laboratory for Computation, Information & Distributed Processing, and an Associate Professor in the Department of Computer Science and in the Genetics, Bioinformatics and Computational Biology Program at Virginia Polytechnic Institute & State University. Dr. Santos has been awarded an NSF CAREER grant, the Spira Award for Excellence in Teaching in Engineering, and the Robinson Faculty Award. She is also currently named to the Top Ten Teacher's List in the College of Engineering at Virginia Tech. She is on the editorial board of Scientific Programming, and subject area editor for the Journal of Supercomputing. She is currently a member of the IDA/DARPA Defense Science Study Group. Reception after the talk in the 2nd Floor Atrium of Siebel Center. |
