NCSA SC05 Demo Abstracts
Grid-Based NAMD Simulation of Gas Transport in Proteins Researchers at the Theoretical and Computational Biophysics Group at the University of Illinois (TCBG) are now able to image gas transport pathways inside proteins, which has many implications for biotechnology and science. For this research, a large number of simulations need to be run and monitored, and the traditional means of doing so is very wasteful of the researchers' time. To solve this problem, a grid submission engine (NAMD-G) targeted towards biomolecular simulations running the NAMD software, has been developed by NCSA in collaboration with TCBG personnel. From the researchers' host machines, NAMD-G can farm out a large number of jobs, in parallel, to the Linux clusters at NCSA and to other supercomputers on the TeraGrid. In addition, NAMD-G monitors and manages multiple sequences of linked simulation jobs, and performs the necessary transfers and backups on all of their associated files on an as-needed basis. During this presentation, the scientific background of the gas transport process, the NIH-supported, freely available (at http://www.ks.uiuc.edu/) NAMD and VMD software, and the NAMD-G submission engine running live, will all be showcased.
Linked Environments for Atmospheric Discovery Linked Environments for Atmospheric Discovery (LEAD) is a NSF funded, large Information Technology Research project aimed at connecting people and technologies with the weather. To this end, LEAD has laid out a project plan whereby we incrementally approach the ultimate goal of seamless interoperability between data sources, forecasting tools, and analysis tools which feed back into data collection schemes through the construction of a number of prototypes with an increasingly large feature set of features relevant to LEAD. In this particular demonstration, we will be showing the materialization of data onto any of our grid and web test bed nodes, with key capabilities in LEAD expressed as services, to be orchestrated through the LEAD portal. This demonstration will show use of the underlying grid to move data as needed between test bed nodes, as well as decoding of the data into the required format for display in the integrated data viewer (IDV).
Computing Katrina: HPC and Hurricanes Katrina was one of the most powerful hurricanes ever to hit the US, and was certainly one of the most devastating. Forecasts of such events, coupled with adaptable emergency response, can greatly reduce casualties and save billions of dollars. We will show how hurricane tracking, coupled with storm surge models, were used to predict how strong, where, and when the storm surge would occur in the hours leading up to Katrina's landfall. Sophisticated visualization techniques will be used to compare these simulations done using our supercomputers with actual observations, including storm surge measurements from sensors, GIS data from aerial mapping and LIDAR, and time-varying multi-spectral satellite images. Our approach will use flow visualization techniques to show time-varying 3D wind fields and storm surges overlayed with large geospatial data from GIS and remote sensing sources . This is an inter-disciplinary effort drawing together various research groups in High Performance Computing, Sensor networks and Visualization to application communities in GIS, Remote Sensing, Coastal and Atmospheric modeling.
Stereoscopic Visualization of Curved Space Black holes are mysterious astrophysical objects which are predicted by Einstein's theory of general relativity. They are supposed to occur in nature when a star collapses and extremely strong gravitational fields form. Einstein's theory describes gravity as the curvature of space and time, a concept which is not intuitively comprehensible from our everyday experience. However, curvature of spacetime results in prominent visual phenomena, which are extremely strong around black holes. By means of computer graphical simulations we can reproduce this theoretical effects. Some of these sound counter-intuitive or even paradoxical when described by words. For instance, the radial distance to the surface of a black hole is very large (getting infinite for a hovering [static] observer), but the circumference is still quite small. Consequently, objects in the vicinity of a black hole look much more distant than objects behind the black hole. Using a High Definition Stereo Display device we try to visually communicate this very behaviour by virtually putting a black hole into an intuitively known environment. The weird effects of the curvature of space become visible. This kind of visualization is presented the first time here at Supercomputing 05 and is still experimental. The computational effort of creating such images is immense and required a supercomputer such as SuperMike at CCT@LSU to performing the rendering.
LTER Pilot GRID: Scaling Environmental Acoustic Monitoring and Analysis Multi-site experiment utilizing Grid infrastructure developed as part of a pilot study for the LTER sites. User selects an audio "signal" they wish to search for in a collection of field captured acoustic data. User selects signal and specific files to search. Files are located, moved to HPC system, Matlab is run for data analysis Pattern Matching, results are sent back to user's browser. Details and visual graphics as well as audio can be produced for each match. Demonstrates a number of Grid services.....
NCSA 2010: Innovative Systems for Science and Engineering Tomorrow's advanced computing systems will include processors with multiple processor cores on each chip, reconfigurable computers based on rapidly advancing field-programmable gate arrays (FPGAs), and heterogeneous systems with fast communications fabrics interconnecting various types of processors. These new systems have the potential to dramatically increase the fidelity and range of simulations and the scope and speed of analysis, data mining and visualization. They also pose significant technical challenges. NCSA Senior Associate Director Rob Pennington will talk about NCSA's plans to be a leader in the instrumentation, measurement, modeling, testing, and evaluation of innovative computing systems -- both hardware and software.
Reconfigurable Systems Applications Programming The Innovative Systems Lab at NCSA is working with application scientists to explore the potential of reconfigurable (FPGA) systems. Insights, lessons learned, and results gained from porting the NAMD, BLAST, and MATPHOT applications to reconfigurable architectures will be presented.
Evolution Highway: Software for Large-Scale Analysis of Mammalian Chromosome Evolution A bioinformatics tool called "Evolution Highway" was developed to visualize the results of a comparative, multi-species mammalian genome analysis. It utilizes the D2K ("Data to Knowledge") environment components created by the Automated Learning Group at the National Center for Supercomputing Applications at University of Illinois at Urbana-Champaign. Evolution Highway is a set of D2K components created to load, correlate and map chromosome and species data to a visual chromosome metaphor for comparative analysis. The D2K framework enables the Evolution Highway to be a desktop application and a web service application (http://evolutionhighway.ncsa.uiuc.edu/). Recently the tool was applied to comparison of genome architectures among eight mammalian species (Murphy et al., Science 309: 613-617, 2005). During the presentation the Evolution Highway interface and features of the program will be introduced. These features include selection of reference genomes for multi-species comparison (currently human or mouse), the alignment of species homologous synteny blocks (HSBs), assignment of centromere and telomere positions, overlay of chromosome HSBs from any of the five reconstructed ancestral genomes and use of user-defined custom tracks. The application of the program to the large-scale comparison of mammalian genome architectures and classification of genome features e.g., evolutionary breakpoints will be discussed.
PRAGMA Grid in Action
NLANR/DAST Network Performance Advisor The NLANR/DAST Network Performance Advisor measures, displays, and analyzes network metrics. Targeted for both end users & network engineers, the Advisor integrates existing diagnostic tools into a common framework; for network engineers & administrators, it provides a customizable interface to monitor network metrics; for end users, Advisor attempts to emulate a junior-level network engineer with its Analysis Engine. The Advisor conforms to the Global Grid Forum's Network Measurement Working Group schema, and thus is able to communicate with third party databases based thereon. We will demonstrate the current capabilities of the Advisor for live reporting, analysis, and archive retrieval. http://dast.nlanr.net/projects/advisor/.
NCSA 2010: Cyberenvironments for Science and Engineering NCSA is creating advanced cyberenvironments to enable scientists and engineers to take full advantage of cyberinfrastructure. Cyberenvironments combine concepts and technologies from distributed/grid computing, digital libraries, collaboratories, problem solving environments, and the semantic web and go beyond an access-to-resources paradigm to provide more natural, end-to-end, science-oriented capabilities. By reducing the effort required to manage large-scale, complex, and interdisciplinary research projects, cyberenvironments will pave the way for enhanced scientific and engineering productivity. NCSA researchers will describe the overall scope of cyberenvironments concept and discuss example capabilities drawn from current projects.
NCSA 2010: Cyber-resources for Science and Engineering The need for high-end computing resources is driven by the increasing fidelity and the attendant sophistication of computational models used to describe natural and engineered systems. Increasingly, it is also driven by the growing quantity of data that must be managed, analyzed, visualized and understood. This talk with give an overview of the more than 40 teraflops of compute power NCSA provides to the research community, education, and industry. It will also show how this hardware infrastructure is complimented by more than 100 scientific and engineering applications and a dedicated 24/7 support staff.
Cybercollaboratory for End-to-End Environmental Research The NCSA Cybercollaboratory is a web-based collaboration system for distributed groups requiring real-time or asynchronous information/data sharing, collaboratively data mining and analysis and other group tasks. We will demonstrate how users in remote locations can easily discuss oil spill and surface current information at Corpus Christi Bay in Texas in real time, using sensor networks and near real-time oil spill trajectory tracking and simulation tools. We will also show how this Cybercollaboratory can provide integrated workflow/dataflow analysis and other capabilities available in the system. This Cybercollaboratory is currently being used as a prototyping cyberenvironment for CLEANER (Collaborative Large-Scale Engineering Analysis Network for Environmental Research) and CUAHSI (Consortium of Universities for Advancement of Hydrologic Science, Incorporated) projects, where NCSA is leading the development for the shared cyberinfrastructure.
GridChem Cyberenvironment for Computational Chemistry In this demonstration we will describe the organization of the Computational Chemistry Grid, a virtual organization to enable computational chemistry community to utilize grid resources. The GridChem client will be demonstrated starting with obtaining an allocation to post-processing their output. The demonstration will include setting personal preferences for resources, job creation (preprocessing), job submission and monitoring, retrieving results and post processing. An automated distribution of jobs using CondorG on existing resources will be attempted. Preferences regarding computers in the grid to use and mass storage device and local disk storage will be available. Job history and accounting for various hierarchical entities will be described. A consulting system for problem and bug reporting and resolution will also be included in the demonstration. The allocation process will be described and the organization will be compared with the existing models for obtaining allocations. A brief discussion of the inner workings and middleware deployed will be included. Future plans for refactoring the architecture and addition of new functionality and applications will be presented. The accounting, middleware server and databases used in the resource discovery will be detailed. Finally, the experience of collaborating in creating cyberenvironments will be highlighted.
iManaging Credentials with MyProxy MyProxy is open source software for managing X.509 Public Key Infrastructure (PKI) security credentials (certificates and private keys). MyProxy combines an online credential repository with an online certificate authority to allow users to securely obtain credentials when and where needed. Users run myproxy-logon to authenticate and obtain credentials, including trusted CA certificates and Certificate Revocation Lists (CRLs). Storing credentials in a MyProxy repository allows users to easily obtain RFC 3820 proxy credentials, without worrying about managing private key and certificate files. They can use MyProxy to delegate credentials to services acting on their behalf (like a grid portal) by storing credentials in the MyProxy repository and sending the MyProxy passphrase to the service. They can also allow trusted servers to renew their proxy credentials using MyProxy, so, for example, long-running jobs don't fail because of expired credentials. A professionally managed MyProxy server can provide a more secure storage location for private keys than typical end-user systems. MyProxy can be configured to encrypt all private keys in the repository with user-chosen passphrases, with server-enforced policies for passphrase quality. By using a proxy credential delegation protocol, MyProxy allows users to obtain proxy credentials when needed without ever transferring private keys over the network. For users that don't already have PKI credentials, the MyProxy Certificate Authority (CA) provides a convenient method for obtaining them. The MyProxy CA issues short-lived session credentials to authenticated users. The repository and CA functionality can be combined in one service or can be used separately. MyProxy provides a set of flexible authentication and authorization mechanisms for controlling access to credentials. Server-wide policies allow the MyProxy administrator to control how credentials may be used. Per-credential policies provide additional controls for credential owners. MyProxy supports multiple authentication mechanisms, including passphrase, certificate, Kerberos, PAM, LDAP, SASL and One Time Passwords (OTP). MyProxy has been used by the grid community for over four years, in projects such as NEESgrid, TeraGrid, EU DataGrid, and the NASA Information Power Grid. MyProxy is included in the NSF Middleware Initiative GRIDS Center software distribution and the Globus Toolkit 4.0 release. In this talk, I'll describe how MyProxy is used in practice today, present new features added in recent MyProxy software releases, and discuss work in progress for providing new MyProxy services with the TeraGrid User Portal.
Managing Astrophysical Simulations with Teuthis Teuthis is a tool to help computational scientists efficiently manage numerical simulations and the data they produce. It is a control panel from which a user can remotely configure and build applications from local source code, then submit jobs based on those applications, track their progress, and transfer the data they produce to other machines. It dramatically reduces the labor required to manage large numbers of independent calculations, allowing parameter studies to be created with a few simple operations. It is also a notebook which organizes calculations so that they may be examined later within the context of the projects that motivated them. They may be assessed, continued, or repeated as desired. The development of Teuthis has been motivated by the requirements of astrophysical simulations on high-performance computers, but it is suited to any computational task that takes a set of input parameters from a file and runs noninteractively. I will demonstrate the use of Teuthis and discuss plans for its public release and future development.
Bioportal: An Extensible Bioinformatics Portal Powered by NCSA and the TeraGrid The North Carolina Bioportal has been designed to be used by both students and researchers. Students and researchers will gain access to over 100 computational tools and biological data sets in a standards compliant environment as well as computational resources. The Bioportal is a shared, extensible bioinformatics portal developed using the Open Grid Computing Environment (OGCE) portal toolkit. NCSA is in the process of deploying the Bioportal for the broader NCSA and TeraGrid communities. The Bioportal includes access to more than one hundred bioinformatics applications, secure access to remote files through GridFTP, desktop and collaboratory tools, such as calendar, bookmarks and news, job management tools, with the ability to monitor jobs and save your results.
Simulated Pore Interactive Computing Environment (SPICE) SPICE aims to understand the vital process of translocation of biomolecules across protein pores by computing the free energy profile of the translocating biomolecule. Without significant advances at the algorithmic, computing and analysis levels, progress on problems of such size and complexity will 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 approaches. SPICE uses sophisticated grid infrastructure to couple distributed high performance simulations, visualization and instruments used in the analysis to the same framework. We describe how we utilize the resources of a federated trans-Atlantic Grid to enhance our understanding of the translocation phenomenon in ways that have not been possible until now.
Open Grid Computing Environments (OGCE) The Open Grid Computing Environments (OGCE) collaboration is an NSF NMI funded project that develops general purpose portlets and services for science gateways. We will provide an overview and demonstration of new OGCE software, including enhanced grid portlets, metadata management, science application management, collaboration portlets using Sakai services, and audio/video collaboration tools. We will also discuss current OGCE work that is concentrating on simplifying the development of portlets from reusable widget components. We will illustrate software using live demos from several OGCE related projects, including the TeraGrid User Portal, NCSA portal work, the LEAD project, and the VLAB computational chemistry portal.
Implementing Effects Due to Stellar Evolution in a Cosmological Simulation Code The evolution of Stars plays an essential role in the formation of galaxies and clusters of galaxies: they light up otherwise invisible dark matter halos, and they inject energy and nuclear reaction products into the interstellar and intracluster media. However, most cosmological simulations cannot directly simulate their formation and evolution because of the vast range of length scales involved -- from millions of parsecs for clusters of galaxies down to 108-pc for an individual star. Hence subgrid models are needed to include their effects in cosmological simulations. Under NCSA's Strategic Applications Program (SAP) we have developed such models for the adaptive mesh refinement (AMR) simulation code FLASH. We will describe the algorithms used in our models and their application to a number of test cases, and we will demonstrate the performance of FLASH on cosmological problems including star formation on NCSA platforms.
|
![[NCSA]](/includes/images/footer_logo.gif){kind=logo}