|
|
|
|
|
|
|
Subscribe to Online or enVision: www.npaci.edu/Press/subscriptions.html
To view the full Online article, append the issue number to the URL: www.npaci.edu/online/v4.x
|
|
NIH's National Institute of General Medical Sciences Awards Funding for Creation of Virtual Cells
SDSC will provide the bioinformatics infrastructure and information coordination environment for a novel scientific endeavor called the Alliance for Cellular Signaling (AFCS), a consortium of scientists at more than 20 academic institutions around the world. AFCS researchers will pursue unsolved biomedical problems, such as communication between the heart muscle cells and immune system cells.
AFCS scientists will communicate by using a sophisticated virtual conferencing system over Internet2. High-capacity computing power will be necessary to house and organize the AFCS database of scientific results; SDSC expertise and resources will provide these capabilities.
The AFCS project is a substantial effort because thousands of signaling molecules have been identified in cells, and assembling a catalog of all of the possible--and more importantly, authentic--interactions among them is a mind-boggling feat. (v4.18)
|
|
|
|
Joint Center for Structural Genomics Funded to Advance High-throughput Protein Structure Determination
To deliver on the promise of the Human Genome Project for health care, scientists must understand the information in the genome, specifically, the functions of the proteins encoded by the genes. The Joint Center for Structural Genomics (JCSG), a consortium led by The Scripps Research Institute, SDSC, and the Stanford Synchrotron Radiation Laboratory at Stanford University has received a National Institutes of Health (NIH) grant of $24 million over a five-year period to expand the knowledge made available in the human and other genome sequencing projects.
The primary aim of the JCSG is to develop and integrate high-throughput robotic technologies to determine the 3-D structure of up to 2,000 proteins, ultimately providing key insights about biological function.
JCSG will develop new structural bioinformatics methods, exploiting SDSC cutting-edge research in the Biology WorkBench, the Protein Data Bank, the SDSC Storage Resource Broker and other activities. The center will also construct new technological systems resulting in a complete, highly automated pipeline within a knowledge-based feedback system. (v4.20)
|
|
|
|
UC San Diego and SDSC Receive NSF ITR Awards to Promote Innovation in Information Technology
NSF has announced the first awards in its Information Technology Research (ITR) initiative for fiscal year 2000. ITR is a new, NSF initiative that aims to promote fundamental research spanning information technology and scientific applications.
One of the leading projects funded through the ITR program is research on large-scale "steerable" software simulations across networks to model the form and function of cells. This $2.5 million project, entitled "Virtual Instruments: Scalable Software Instruments for the Grid" is led by Francine Berman, professor of computer science and engineering at the Jacobs School of Engineering and an SDSC fellow.
UC San Diego and SDSC were also awarded $625,000 as part of the $11.8 million Grid Physics Network project led by the Universities of Chicago and Florida. Technology being developed within NPACI by SDSC's DICE group is directly applicable to many of the virtual data grid requirements, including data-handling systems, information discovery services, and the ability to execute remote proxies. DICE researchers will also collaborate with Keith Marzullo, professor of computer science and engineering at the Jacobs School of Engineering to integrate UC San Diego-developed technologies into this grid environment. (v4.19)
|
|
|
|
Latest HSI Developments from SDSC Improve Performance, Usability of Upgraded HPSS at Caltech, SDSC
Researchers who store simulation results
on the High-Performance Storage System (HPSS) at NPACI partner sites can now achieve better performance during data transfers using the most recent versions of the Hierarchical Storage Interface (HSI).
The most significant enhancement is a rewritten I/O component that allows file transfers to achieve 90% to 95% of the maximum rate permitted by the slowest link on the end-to-end transfer path between HPSS and the HSI client. In practice, most users will see transfer rates five or six times faster than previous versions of HSI.
In addition, Caltech's Center for Advanced Computing Research has completed a software and hardware upgrade of its HPSS installation to improve access rates and system stability. At SDSC, 16 tape drives were upgraded on the center's HPSS, doubling the maximum capacity to 360 TB. SDSC continues to operate the world's largest production installation of HPSS, with more than 200 terabytes of data stored.
HSI is fast becoming the de facto standard at all HPSS installations, and for users of NPACI's high-performance systems at SDSC, the I/O-enhanced HSI is now the default file transfer client. (v4.19)
|
|
|
|
SDSC Volume Explorer Rendering Software Released
SDSC has released SDSC Volume Explorer, a program that allows a user to interactively render and explore volume data sets in real time on a desktop computer configured with a Mitsubishi VolumePro 500 graphics card. The new software package--downloadable at no charge from the Web--and the relatively inexpensive graphics card give high-powered graphics performance to desktop computers running the Sun Solaris, SGI Irix, and Microsoft Windows operating systems.
Applications in which volume information is important include medical imaging, computational fluid dynamics, stress analysis, meteorology, oceanography, and astronomy. Being able to interactively change the transfer function gives new insights into the data, especially for cases in which the information for nearby voxels obscures the view deeper into the data volume.
Due to the need to blend contributions from all voxels along a line of sight according to their transfer functions, direct volume rendering has always been slow to do interactively. SDSC Volume Explorer performs 500 million ray-trace voxel composites per second, providing real-time direct display of volume data. (v4.18)
dvl.sdsc.edu/vx
|
|
|
|
Demonstrations Prove Capabilities of Silver Metascheduler
Using resources located in the Molecular Science Computing Facility (MSCF) at the Pacific Northwest National Laboratory and SDSC, a number of metascheduling demonstrations were performed on July 8, 2000, and reported at the 4th Grid Forum Workshop July 10-12, 2000. The demonstrations featured NPACI's 1,152-processor Blue Horizon and the MSCF's 96-processor NWecs1 system, both IBM SP systems
Silver is a metascheduling computer system that improves system efficiency by using the aggregate resources of multiple high-performance systems while allowing local schedulers to optimize local resources.
The Silver demonstration managed jobs across heterogeneous architectures, in multiple administrative domains with non-uniform user and file spaces, which necessitates specifying job requirements and attributes on a per-system basis. In essence, it proved the ability to co-allocate resources of various types across multiple systems, and to execute jobs requiring resources larger than what could be found at any single site. (v4.20)
|
|
|
|
NSF Director Highlights SDSC EOT Contributions during Conference Keynote
Among the highlights of the September
14-16 Grace Murray Hopper Celebration of Women in Computing 2000 conference, held in Hyannis, Massachusetts, was a keynote address delivered by NSF Director Rita Colwell that highlighted the contributions of SDSC scientists Rozeanne Steckler and Mike Bailey toward attracting more girls and women to computing and science.
Colwell spoke about the need to understand why more women are not in computing careers. As Colwell held up a Girl Scout computer badge, she described the Girl Scout USA and SDSC Science Interest Group that has been led by Steckler and Bailey for the past several years.
"Rozeanne and Mike's projects are clearly making a difference in the bigger picture of computer science education and outreach, and we feel fortunate to have them at SDSC," said SDSC EOT Director Ann Redelfs. "They play a tremendous role in our outreach efforts to women and to other underrepresented groups as well." (v4.19)
|
|
|
|
UC San Diego Wireless Network Project Collaborates with Pala Band of Indians for Education Opportunities
A collaboration between the Pala Band of
Indians and the UC San Diego High-Performance Wireless Research and Education Network (HPWREN) research project funded by NSF is developing educational programs to take advantage of high-speed wireless Internet service to the tribe's remote reservation.
One of the ways in which the Pala people are using the high-speed Internet connectivity provided by HPWREN is to educate their children. Doretta Musick, coordinator of the Pala Learning Center, is working closely with SDSC to develop courses that will allow the students to make good use of the Internet and foster their educational objectives. (v4.21)
hpwren.ucsd.edu
|
|
|
|
NPACI Participants Present at SIAM First Conference on Computational Science and Engineering
Several NPACI researchers spoke at the Society for Industrial and Applied Math (SIAM) First Conference on Computational Science and Engineering, which was held September 21-24 in Washington, D.C. The meeting focused on the wide array of major computational efforts on large problems in science and engineering.
Kris Stewart of San Diego State University (SDSU), director of NPACI's Education Center on Computational Science and Engineering, spoke on computational science education at two minisymposia. Jose Castillo of SDSU and Steve Cunningham of California State University, Stanislaus, also joined Stewart in another minisymposium on "Computer Graphics and the Continuum of Computational Science Education."
Additional NPACI community speakers included Klaus Schulten (University of Illinois), Chandrajit Bajaj (University of Texas, Austin), and Mary Wheeler (University of Texas, Austin). (v4.20)
|
|
|
|
|