HPC Infrastructure II             Event Type Start Time End Time Rm # Chair     Masterworks 3:30PM 4:15PM 16-18 Beverly Clayton (Pittsburgh Supercomputing Center)   Title: Informatics and Pharmacogenomics   Speakers/Presenter: John Reynders (Vice-President, Informatics, Celera)   Masterworks 4:15PM 5:00PM 16-18 Beverly Clayton (Pittsburgh Supercomputing Center)   Title: The Plan for BlueGene/L Simulation Environment: Roadmap for Enabling Scientific Simulation at Unprecedented Scale   Speakers/Presenter: Mark Seager (Integrated Computing and Communication Asst. Dept. Head for Advanced Technology, Lawrence Livermore National Laboratory)

Session: HPC Infrastructure II   Title: Informatics and Pharmacogenomics   Chair: Beverly Clayton (Pittsburgh Supercomputing Center)   Time: Tuesday, November 18, 3:30PM - 4:15PM   Rm #: 16-18   Speaker(s)/Author(s):     John Reynders (Vice-President, Informatics, Celera)       Description:   Celera's Informatics environment has transitioned from the singular effort of assembling the human genome to a broad range of analytic, computational, and knowledge-based efforts focused upon the research and development of pharmacogenomic therapeutics. This presentation will discuss how Celera's Informatics programs and platforms build upon Celera's heritage in simulation, software, and bioinformatics to integrate and interpret large-scale and heterogeneous genetic, proteomic, and laboratory data from experimental and in-silico data sources to discover and develop novel pharmacogenomic therapeutics.   Link: --

Session: HPC Infrastructure II   Title: The Plan for BlueGene/L Simulation Environment: Roadmap for Enabling Scientific Simulation at Unprecedented Scale   Chair: Beverly Clayton (Pittsburgh Supercomputing Center)   Time: Tuesday, November 18, 4:15PM - 5:00PM   Rm #: 16-18   Speaker(s)/Author(s):     Mark Seager (Integrated Computing and Communication Asst. Dept. Head for Advanced Technology, Lawrence Livermore National Laboratory)       Description:   With the advent of IBM's cellular approach to building large scale parallel platforms, it becomes possible to consider the ramifications of a machine with 65,536 nodes 131,072 CPUs capable of a peak 360 teraFLOP/s. The scale of this machine is unprecedented and the science that could be accomplished is also astonishing. However, several major challenges remain to the utilization of this platform in a scientific simulation environment. In this talk we outline the key issues: scalability of the hardware, scalability of the system software, scalability of the file system, scalability of applications and scalability of the support infrastructure. We also outline our plan for addressing these scalability challenges. We discuss the simple is better hardware scalabiliity strategy, the divide and conquer strategy for system software and file system strategy. From the file system strategy we describe the full simulation environment (including parallel visualization and hierarchical storage management). Finally we outline the plan for scaling up real scientific applications for BlueGene/L.   Link: --