BEGIN:VCALENDAR VERSION:2.0 X-WR-CALNAME:PSC Live! at SC|05 PRODID:-//Apple Computer\, Inc//iCal 2.0//EN X-WR-RELCALID:44126FA5-AF43-4EE6-9570-70EABBDC4492 X-WR-TIMEZONE:US/Eastern CALSCALE:GREGORIAN METHOD:PUBLISH BEGIN:VTIMEZONE TZID:US/Pacific LAST-MODIFIED:20051109T192041Z BEGIN:STANDARD DTSTART:20051030T090000 TZOFFSETTO:-0800 TZOFFSETFROM:+0000 TZNAME:PST END:STANDARD BEGIN:DAYLIGHT DTSTART:20060402T010000 TZOFFSETTO:-0700 TZOFFSETFROM:-0800 TZNAME:PDT END:DAYLIGHT END:VTIMEZONE BEGIN:VEVENT LOCATION:Booth 722 DTSTAMP:20051108T173311Z UID:0755AE39-DED5-4873-9DB0-E6E538928992 SEQUENCE:4 DTSTART;TZID=US/Pacific:20051116T150000 SUMMARY:A Case Study in Combining Bioinformatics and Computational Chemi stry Analyses of an Enzyme Family DTEND;TZID=US/Pacific:20051116T160000 DESCRIPTION:by Hugh Nicholas\, Pittsburgh Supercomputing Center \n\nWe d escribe an integrated combination of bioinformatics\, molecular dynamics \, and quantum mechanics calculations on the Aldehyde Dehydrogenase enzy me family that provides a comprehensive characterization of that enzyme family illuminates many open questions and incompletely understood exper imental observations about that family. A comprehensive bio-informatics analysis of the Aldehyde Dehydrogenase enzyme superfamily was used to id entify the sequence and structural elements that are essential to the fu nctioning of all of the enzymes in the superfamily\, as well as those se quence and structural elements critical to individual families. These es sential and critical elements are described and integrated with structur al and experi-mental information which serves as the basis for mixed mol ecular and quantum mechanical calculations that serve to define the geom etry of substrate binding and elucidate the steps in the catalytic mecha nism. These calculations show a unique catalysic mechanism\, which will be described. These calculations indicate that distinct genetic diseases involving different mutations in different Aldehyde Dehydrogenases appa rently have the same mechanistic basis.\n END:VEVENT BEGIN:VEVENT LOCATION:Booth 722 DTSTAMP:20051108T172958Z UID:90EC69B9-758C-4FE6-9A70-3EB92B21CE23 SEQUENCE:6 DTSTART;TZID=US/Pacific:20051115T143000 SUMMARY:Dynamic Data-driven Inversion for Terascale Simulations: Real-ti me Identification of Airborne Contaminants DTEND;TZID=US/Pacific:20051115T151500 DESCRIPTION:by George Biros\, University of Pennsylvania\n\nIn contrast to traditional terascale simulations that have known\, fixed data inputs \, dynamic data-driven (DDD) applications are characterized by unknown d ata and informed by dynamic observations. DDD simulations give rise to i nverse problems of determining unknown data from sparse observations. Th e main difficulty is that the optimality system is a boundary value prob lem in 4D space-time\, even though the forward simulation is an initial value problem. We construct special-purpose parallel multigrid algorithm s that exploit the spectral structure of the inverse operator. Experimen ts on problems of localizing airborne contaminant release from sparse ob servations in a regional atmospheric transport model demonstrate that 17 -million-parameter inversion can be effected at a cost of just 18 forwar d simulations with high parallel efficiency. On 1\,024 Alphaserver EV68 processors\, the turnaround time is just 29 minutes. Moreover\, inverse problems with 135 million parameters corresponding to 139 billion total space-time unknowns are solved in less than 5 hours on the same number of processors. These results suggest that ultra-high resolution data-dri ven inversion can be carried out sufficiently rapidly for simulation-bas ed real-time hazard assessment.\n END:VEVENT BEGIN:VEVENT DURATION:PT1H LOCATION:Booth 722 DTSTAMP:20051108T175118Z UID:F9EB5C7E-F3F4-491A-8931-FE967ABFDFBD SEQUENCE:4 DTSTART;TZID=US/Pacific:20051115T120000 SUMMARY:Production Science on PSC's\nCray XT3 DESCRIPTION:by Nick Nystrom\, Pittsburgh Supercomputing Center\n\nBigBen \, the Pittsburgh Supercomputing Center's Cray XT3\,\n entered productio n as an NSF TeraGrid resource on October 1. \n Users are now running a wide variety of applications\, for\n example end-to-end earthquake simu lation and achieving more\n than 8 teraflops modeling electronic structu re in materials\n science. In this talk\, we will survey applications r unning\n on PSC's Cray XT3\, emphasizing both scientific results and\n p erformance. END:VEVENT BEGIN:VEVENT DURATION:PT1H LOCATION:Booth 722 DTSTAMP:20051108T173324Z UID:7CE08281-A900-4ECE-BA64-0EF2FB431298 SEQUENCE:2 DTSTART;TZID=US/Pacific:20051116T160000 SUMMARY:Interactive Control of Remote Supercomputer Simulations DESCRIPTION:by Paul Woodward\, University of Minnesota\n\nThe ability to quickly do short\, exploratory runs of PPM simulations of turbulence wi ll have a significant impact on scientific productivity. The Cray XT3 ma chine at Pittsburgh Supercomputing Center has the potential to compute a relatively small problem\, on a grid of just 512^3 cells\, fast. Previo us systems are only able to achieve their best performance on extremely large problems. LCSE demonstrates a prototype computational steering\, v isualization and data analysis system that will be able to produce volum e-rendered images from this data at a rate of a few per frames second. I n this demonstration visualization data will be streamed from the Cray X T3 in Pittsburgh directly to the show floor\, and the visualization and the computation will be interactively steered from the show floor via th e TeraGrid backbone network. END:VEVENT BEGIN:VEVENT DURATION:PT1H LOCATION:Booth 722 DTSTAMP:20051109T192026Z UID:10F2067C-C6F8-4711-A38F-B7EB0B705C4A SEQUENCE:7 DTSTART;TZID=US/Pacific:20051116T170000 SUMMARY:Scalable Simulation of the Heart and Other Organs DESCRIPTION:by Katherine Yelick\, UC Berkeley and Lawrence Berkeley Nati onal Laboratory\n\nSimulation is often called the \"third pillar of scie nce\,\" along with theory\nand experimentation. Simulation of the human body would enable a virtual\nexperimental setup that would have applica tions in biology and medicine.\nWhile a full simulation of the human bod y is far from possible today\,\nindividual models exist of many of the o rgans within the body. One class of\nproblems that arise in such simula tions is the modeling of fluid flow within\nan organ\, often when that f luid contains immersed elastic structures such as\nmuscle\, membrane\, o r other tissue. The computational cost of modeling the\nfluid dynamics even within a single organ is very high\, requiring the use of\ntoday's fastest parallel machines.\n\nIn this talk I will describe a scalable pa rallel algorithm for the immersed\nboundary method. The method\, due to Peskin and McQueen\, has been used to\nsimulate blood flow in the heart \, blood clotting\, the motion of bacteria and\nsperm\, embryo growth\, and the response of the cochlea to sound waves. Our\nparallel implement ation uses a novel programming language called Titanium\,\nwhich is a hi gh performance extension of Java. I will describe the Titanium\nlanguag e and compiler as well as our computational framework for the\nimmersed boundary method\, which is designed to be extensible and is publicly\nav ailable along with the Titanium compiler. I will also talk about some o f\nthe remaining open problems in Computer Science and Applied Mathemati cs\nmotivated by this application domain.\n\nThis work was done in colla boration with Ed Givelberg\, Armando Solar\,\nCharles Peskin and Dave Mc Queen.\n END:VEVENT BEGIN:VEVENT LOCATION:Booth 722 DTSTAMP:20051108T172916Z UID:F90F57E6-80B5-4E5C-A380-412C1499753A SEQUENCE:5 DTSTART;TZID=US/Pacific:20051116T100000 SUMMARY:TeraGrid Cross-Site Simulations and Visualizations of the Human Arterial Tree DTEND;TZID=US/Pacific:20051116T110000 DESCRIPTION:by George Karniadakis\, Brown University\n\nCardiovascular d isease accounts for almost fifty percent of deaths in the western world. The formation of arterial disease\, such as atherosclerotic plaques\, i s strongly related to the blood flow patterns\, and is observed to occur preferentially in regions of separated and re-circulating flow such as vessel branches and bifurcations. In this talk we will perform for first time simulations of blood flow in the entire human arterial tree throug h detailed three-dimensional computations at a number of arterial bifurc ations\, coupled by the wave-like nature of pulse information traveling from the heart to arteries that is modeled by a reduced set of one-dimen sional equations. We employ MPICH-G2 and conduct geographically-distribu ted coupled cross-site simulations at major TeraGrid sites in the US and high-end systems in the UK. Flow visualizations on all arteries simulta neously will also be demonstrated using the TeraGrid resources.\n END:VEVENT BEGIN:VEVENT DURATION:PT1H LOCATION:Booth 722 DTSTAMP:20051108T173906Z UID:5516EFC6-87F7-4FF5-A4E1-87E8D3F04E17 SEQUENCE:5 DTSTART;TZID=US/Pacific:20051117T130000 SUMMARY:Simulating Enzyme Reactions on Supercomputers: Hybrid Transfer i n ALDH DESCRIPTION:by Shawn Brown\, Pittsburgh Supercomputing Center\n \nWe per formed large scale simulations of Aldehyde Dehydrogenase chemistry on th e Cray XT3 using a hybrid Quantum Mechanical/Molecular Mechanical (QM/MM ) potential energy function. The simulations used a software package ca lled DYNAMO and performed over 900 processors. This enabled us to exami ne a large area of critical atomic interactions that govern a crucial st ep in ALDH catalysis. Genetic mutations that cause changes in these int eractions result in two known metabolic diseases\, Sjogren-Larsson Syndr ome and Type II Hyperprolinemia. END:VEVENT BEGIN:VEVENT LOCATION:Booth 722 DTSTAMP:20051108T173157Z UID:9647531B-60F6-4E2F-8A09-E52CC1BA955E SEQUENCE:5 DTSTART;TZID=US/Pacific:20051116T110000 SUMMARY:VORTONICS: Grid Computing for the Simulation and Identification of Vortical \nStructures in Fluids\n DTEND;TZID=US/Pacific:20051116T120000 DESCRIPTION:by Bruce Boghosian\, Tufts University\n\nVortex knots and li nks are the most evident coherent structures of fluid turbulence. Like t he elementary particles of high-energy physics\, there are a wide variet y of such structures\; more exotic ones are found at higher energies\, a nd they have a characteristic lifetime after which they decay into other such structures. An improved understanding of the creation\, behavior\, and interaction of vortical structures would elucidate fundamental ques tions in theoretical fluid dynamics\, such as the reconnection problem a nd the detachment of “hairpin” vortices from boundaries in the intermitt ency approach to turbulence. It holds the potential for improved underst anding in applied fluid dynamics\, including the dynamics of meteorologi cal structures such as tornadoes and hurricanes. It may also lead to imp rovements in computational fluid dynamics\, such as naturally adaptive v ortex algorithms that avoid vortex tangles and requisite “vortex surgery .” High-resolution direct numerical simulations of the dynamics of vorti cal structures at high Reynolds number are\, however\, among the most in tensive scientific computations attempted today. \n\nWe describe the VOR TONICS package for locating and tracking vortex cores\, and addressing o ther fundamental problems of topological fluid dynamics. The package inc ludes modules for generating\, evolving\, and identifying vortex cores a nd routines for Fourier-resizing and remapping the computational lattice . It is parallelized using MPI and geographically distributed using MPIC H-G2. It also includes a methodology for computational steering\, checkp ointing\, rewinding\, and dynamic adjustment of resolution. We describe the geographical distribution of the tasks performed by this code\, as w ell of the domain-decomposition of the computational lattice itself on t he TeraGrid and other distributed HPC resources.\n END:VEVENT BEGIN:VEVENT DURATION:PT1H LOCATION:Booth 722 DTSTAMP:20051108T173534Z UID:64E9126C-C743-41A5-B78B-4A4E05CC0EE3 SEQUENCE:5 DTSTART;TZID=US/Pacific:20051117T120000 SUMMARY:Teraflop Computing for Nanoscience DESCRIPTION:by Yang Wang\, Pittsburgh Supercomputing Center \n\nOver the last two-three decades there has been significant progress in the first principles methods to calculate the properties of materials at the quan tum mechanics level. These methods have largely been based on the local density approximation (LDA) to density functional theory (DFT). However\ , nanoscience places new demands on these first principles methods becau se of the large numbers of atoms\, in the range of thousands to millions of atoms\, present in even the simplest of nano-structured materials. F ortunately\, recent advances in the locally self-consistent multiple sca ttering (LSMS) method are making the direct quantum mechanical simulatio n of nano-structured materials possible. The LSMS method is an order-N a pproach to the first principles electronic structure calculation. It is highly scalable on massively parallel processing super-computers and is best suited for performing large unit cell simulations to study the elec tronic and magnetic properties of materials with complex structure. In t his presentation\, I will give a brief introduction of the LSMS method a nd will show that this method\, aided with the state of the art teraflop computing technology\, effectively accomplishes the first step towards understanding the electronic and magnetic structure of nanoparticles wit h dimension size up to 10 nanometers. \n\nI will demonstrate\, as an exa mple\, the electronic and magnetic structure calculated for an iron nano particle embedded in iron aluminide crystal matrix. I will also explain to what extent future petaflop computing systems may enable the realisti c quantum mechanical simulation of real nano-structured materials.\n END:VEVENT BEGIN:VEVENT LOCATION:Booth 722 DTSTAMP:20051108T173451Z UID:3D665F98-D45E-4AED-BAB7-79094E18F3C7 SEQUENCE:4 DTSTART;TZID=US/Pacific:20051117T100000 SUMMARY:Computational Cosmology and Cosmogony DTEND;TZID=US/Pacific:20051117T110000 DESCRIPTION:by Thomas Quinn\, University of Washington\n\nThe dynamics o f self gravitating systems play a significant role over a wide range on length scales in Astrophysics. On the largest scales\, connecting the f luctuations in the Cosmic Microwave Background to the galaxies we observ e today requires following gravitational collapse for over 10 orders of magnitude in density. Furthermore\, the physics of gas dynamics and sta r formation must be accurately modeled in order to fully make the connec tion between \"mass\" and \"light\". On the other end of the size spect rum\, protoplanetary systems must be followed for millions of years to u nderstand the formation history of planets.\n\nI will describe how simul ations on the PSC supercomputers have contributed to our understanding o f planet formation and galaxy formation. I will also discuss some of th e computational challenges we still face in these disciplines and strate gies for tackling them. \n END:VEVENT BEGIN:VEVENT LOCATION:Booth 722 DTSTAMP:20051108T172952Z UID:8085FE24-4BFB-4AAF-B0D8-98936AC53508 SEQUENCE:5 DTSTART;TZID=US/Pacific:20051115T110000 SUMMARY:Simulated Pore Interactive Computing Environment (SPICE): Using Grid Computing to Understand DNA Translocation Across Protein Nanopores Embedded in Lipid Membranes DTEND;TZID=US/Pacific:20051115T120000 DESCRIPTION:by Peter Coveney\, University College London\n\nSPICE aims t o understand the vital process of translocation of biomolecules across p rotein pores by computing the free energy profile of the translocating b iomolecule. Without significant advances at the algorithmic\, computing and analysis levels\, progress on problems of such size and complexity w ill remain beyond the scope of computational science for the foreseeable future. Grid computing provides the required new computing paradigm as well as facilitating the adoption of new algorithmic and analytical appr oaches. SPICE uses sophisticated grid infra-structure to couple distribu ted high performance simulations\, visualization and instruments used in the analysis to the same framework. We describe how we utilize the reso urces of a federated trans-Atlantic Grid to enhance our understanding of the translocation phenomenon in ways that have not been possible until now.\n END:VEVENT BEGIN:VEVENT DURATION:PT1H LOCATION:Booth 722 DTSTAMP:20051108T173536Z UID:D4F3BC26-6302-4DEC-A747-E475FAD7E26C SEQUENCE:2 DTSTART;TZID=US/Pacific:20051117T110000 SUMMARY:Interactive Control of Remote Supercomputer Simulations DESCRIPTION:by Paul Woodward\, University of Minnesota\n\nThe ability to quickly do short\, exploratory runs of PPM simulations of turbulence wi ll have a significant impact on scientific productivity. The Cray XT3 ma chine at Pittsburgh Supercomputing Center has the potential to compute a relatively small problem\, on a grid of just 512^3 cells\, fast. Previo us systems are only able to achieve their best performance on extremely large problems. LCSE demonstrates a prototype computational steering\, v isualization and data analysis system that will be able to produce volum e-rendered images from this data at a rate of a few per frames second. I n this demonstration visualization data will be streamed from the Cray X T3 in Pittsburgh directly to the show floor\, and the visualization and the computation will be interactively steered from the show floor via th e TeraGrid backbone network. END:VEVENT BEGIN:VEVENT LOCATION:Booth 722 DTSTAMP:20051108T173152Z UID:8B025A5E-216F-4B0E-B531-73EEE58DFC9E SEQUENCE:4 DTSTART;TZID=US/Pacific:20051116T130000 SUMMARY:HPN-SSH: Easy\, Fast\, and Secure Communication in HPN Environme nts DTEND;TZID=US/Pacific:20051116T140000 DESCRIPTION:by Chris Rapier\, Pittsburgh Supercomputing Center \n\nData transfer across high performance networks usually means being given a me nu of fast\, easy\, and secure and only being able to pick two. HPN-SSH changes that by giving even novice users access to high speed\, cryptogr aphically secure\, and easy to use data transfers. Speeds of more than 6 00 Mb/s have been documented. Additionally\, HPN-SSH includes a new port forwarding utility that will automatically and transparently create sec ure communication tunnels for any sort of connection and in any environm ent. This allows for ubiquitous security for mobile users.\n END:VEVENT BEGIN:VEVENT LOCATION:Booth 722 DTSTAMP:20051108T173239Z UID:FD4F8AEE-AFFE-4A0C-9012-522B8A07303E SEQUENCE:4 DTSTART;TZID=US/Pacific:20051116T140000 SUMMARY:Simulating SIMOX: Joint Grid Experiments on the Japan-USC-TeraGr id Testbed DTEND;TZID=US/Pacific:20051116T150000 DESCRIPTION:by Yoshio Tanaka\, AIST\n\nA large-scale and long-running si mulation of Separation by Implantation of Oxygen for creating a semicond uctor structure was gridified and executed using both GridRPC and MPI. T he experiment shows that the new programming approach enables applicatio ns to be (1) flexible: allow dynamic resource allocation/migration\; (2) robust: detect errors and recover from faults automatically for long ru ns\, and (3) efficient: manage thousands of CPUs distributed on an inter continental scale.\n END:VEVENT BEGIN:VEVENT LOCATION:Booth 722 DTSTAMP:20051108T172955Z UID:7496FAF0-9EF0-45C1-9DF6-023B2A8AB7B2 SEQUENCE:6 DTSTART;TZID=US/Pacific:20051115T133000 SUMMARY:Large Scale Inverse Earthquake Modeling on Terascale Supercomput ers DTEND;TZID=US/Pacific:20051115T143000 DESCRIPTION:by Omar Ghattas\, University of Texas at Austin\n\nOptimizat ion of systems governed by large-scale PDE simulations presents manifold challenges and opportunities. We begin with illustrations of some drivi ng applications involving contaminant transport\, cardiac mechanics\, ar tificial heart design\, and design of linear accelerators. We then focus on a challenging PDE-constrained optimization problem arising in invers e earthquake modeling: that of estimating the earth model in earthquake simulations from observations of ground motion from past earthquakes. We discuss such compounding issues as extreme large scale\, ill-posedness\ , discontinuous solutions\, challenges in constructing preconditioners f or the reduced Hessian\, nonlinear convergence difficulties\, and multip le local minima\, and techniques for addressing them. We end with a disc ussion of some outstanding challenges and the conclusion that earthquake inversion to frequencies of engineering interest remains a major challe nge for petascale computing.\n END:VEVENT BEGIN:VEVENT LOCATION:Booth 722 DTSTAMP:20051108T172946Z UID:004DA912-B63E-4FD8-A949-DE948F7C7FFD SEQUENCE:6 DTSTART;TZID=US/Pacific:20051115T100000 SUMMARY:Interactive Control of Remote Supercomputer Simulations DTEND;TZID=US/Pacific:20051115T110000 DESCRIPTION:by Paul Woodward\, University of Minnesota\n\nThe ability to quickly do short\, exploratory runs of PPM simulations of turbulence wi ll have a significant impact on scientific productivity. The Cray XT3 ma chine at Pittsburgh Supercomputing Center has the potential to compute a relatively small problem\, on a grid of just 512^3 cells\, fast. Previo us systems are only able to achieve their best performance on extremely large problems. LCSE demonstrates a prototype computational steering\, v isualization and data analysis system that will be able to produce volum e-rendered images from this data at a rate of a few per frames second. I n this demonstration visualization data will be streamed from the Cray X T3 in Pittsburgh directly to the show floor\, and the visualization and the computation will be interactively steered from the show floor via th e TeraGrid backbone network. END:VEVENT BEGIN:VEVENT LOCATION:Booth 722 DTSTAMP:20051108T173002Z UID:384B5295-9F22-4EE8-AC79-8F7D0936B699 SEQUENCE:5 DTSTART;TZID=US/Pacific:20051115T153000 SUMMARY:From Physical Modeling to Scientific Understanding: An End-to-En d Approach\nto Parallel Supercomputing\n DTEND;TZID=US/Pacific:20051115T161500 DESCRIPTION:by Tiankai Tu\, Carnegie Mellon University\n\nTraditional hi gh-performance parallel scientific computing adopts an offline approach where files are used as interface between simulation components\, such a s meshing\, partitioning\, solving and visualizing. Unfortunately\, such an approach results in time-consuming file transfers\, disk I/O and dat a format conversions that consume large amounts of network\, storage\, a nd computing resources while contributing nothing to applications. We pr opose an end-to-end approach to parallel supercomputing. The key idea is to replace the cumbersome file interface with a scalable\, parallel\, r untime data structure\, on top of which all simulation components are co nstructed in a tightly coupled way. We have implemented our new methodol ogy within an octree-based finite element simulation system named Hercul es. The only input to Hercules is material property descriptions of a pr oblem domain\; the only outputs are small jpeg-formatted images generate d as they are simulated at every visualization time step. There is absol utely no other intermediary file I/O. Performance evaluation of Hercules on up to 2\,048 processors on the Cray XT3 system and the AlphaServer s ystem at Pittsburgh Supercomputing Center has shown good isogranular sca lability and fixed-size scalability.\n END:VEVENT END:VCALENDAR