REDMOND, Wash. — June 11, 2008 — While many students leave campus for summer vacation, university research teams across the country are gearing up to address some of the world’s most challenging technical computing problems with Microsoft Corp.’s high-performance computing (HPC) solutions. Eight of the nation’s top supercomputing institutions — Louisiana State University; Rutgers, The State University of New Jersey; University of Arizona; University of Florida; University of Iowa; University of Nebraska at Omaha; University of North Carolina at Charlotte; and University of Washington — now running Windows Compute Cluster Server 2003 (CCS 2003) have chosen Microsoft to undertake ventures that will bring benefits to their research, teaching and learning process.
“With this easy-to-use platform, these universities are efficiently and cost-effectively developing and deploying the HPC systems their talented researchers need and expect,” said Anthony Salcito, general manager, U.S. Enterprise Education at Microsoft. “Our Microsoft team is working side-by-side with universities to make advanced server technologies available to university researchers who are creating solutions to some of society’s greatest challenges.”
Researchers moving to CCS 2003 report that it yields performance gains and new capabilities even without any optimization. Valerie Daggett, University of Washington professor and principal investigator of the Daggett Research Group within the university’s Department of Bioengineering, reported that the simulation software performed 5 percent better on CCS 2003 than on the previous heavily optimized version that was running on identical hardware. CCS 2003 also enabled Daggett to double-task a cluster, using it for both simulation and analysis.
Projects Showcase HPC Impact in Diverse Areas of Study
The U.S. universities working on the Microsoft platform vary in size, scope and focus. Projects include the following:
The world’s largest Windows-based cluster computing center went online this year at the Holland Computing Center (HCC) at the University of Nebraska at Omaha. In addition to enabling research among the university’s faculty, students and researchers from around the country, the 1,150-node cluster supports the Department of Defense, Gallup Inc., Microsoft and the actuarial consulting firm Milliman Inc. The cluster also supports genetics researchers at Omaha’s Henry Doorly Zoo and researchers studying protein strands at the University of Nebraska Medical Center. More information about HCC is available at http://hollandhpc.com/index.shtml.
The University of Florida, University of Arizona and Rutgers have joined forces to establish a national research Center for Autonomic Computing (CAC). The collaborative group is funded by the National Science Foundation (NSF) and combines resources from the three universities, private companies and the federal government to make a variety of computer systems and applications. Their efforts include making everything from desktop computers to sophisticated air traffic control systems reliable, secure, efficient and easy to manage. The center uses CCS 2003 to generate the autonomic computing system that is designed to function with minimal management even as conditions and users change.
“Autonomic computing algorithms can greatly reduce the growing costs of administrating computer systems and help protect against loss of service in systems performing critical functions, including those managing power grids, stock markets and hospital networks,” said José Fortes, director of the new center at the CAC’s University of Florida site. “They can also greatly improve the speed and efficiency of complex systems that utilize a large number of hardware and software components.”
More information about the CAC is available at http://nsfcac.org/index.html.
At Rutgers, CCS 2003 is being used as a platform for students to develop and experiment with autonomic concepts and technologies for managing datacenter systems and applications. For example, students are building financial and biomedical informatics applications on the platform to support dynamic scale-out and failure resilience. Rutgers is also looking into others ways that it can use Microsoft’s HPC solutions.
“We are looking forward to moving to Windows HPC Server 2008 when it becomes available specifically for research at the Center for Autonomic Computing at Rutgers,” said Manish Parashar, professor at Rutgers and co-director of the CAC. “HPC Server 2008 will provide us with some key capabilities such as integrated virtualization support, which we can use to support a wide class of applications. It will also provide us with interesting autonomic behaviors for power and energy management, performance and productivity management, and dynamic on-demand scaling.”
Research generated from HPC at the University of Arizona Autonomic Computing Lab site ranges from helping scientists study global climate change to accelerating the sharing of knowledge across the pharmaceutical industry. This fall, University of Arizona scientists will begin to look at how plants acquire and allocate resources to survive, grow and reproduce; how global climate change is affecting those processes; and how those changes impact the larger biological ecosystem. For example, scientists will be able to understand how plants will use water in future climates, which could determine how much water will be available for use by society.
“The integration of the plant models, the experimental environment, and visualization and analysis tools at runtime can be efficiently supported using Microsoft CCS tools and services,” said Salim Hariri, director of the CAC’s University of Arizona site.
The Daggett Research Group within the University of Washington’s Department of Bioengineering is world-renowned in the study of protein stability, function and folding — some of the fundamental unsolved problems in molecular biology. Although much is known of native folded conformation of proteins, very little is known about the actual folding process, the understanding of which has important implications for research into all biological processes, including aging and human diseases.
Experimental approaches provide only limited amounts of information, so the Daggett Research Group uses computer simulations that require massive amounts of computation and the need for more computing power. The group augmented its existing HPC resources with two new HPC systems based on CCS 2003, which has enabled the group to make fundamental breakthroughs in how the results of simulations are analyzed.
“With the combination of Windows Compute Cluster Server 2003 and Microsoft SQL Server 2005, we can attack problems in new ways and at new magnitudes,” Daggett said. “We can examine 100 times more data, because tasks that used to take hours are now reduced to fractions of a second. It’s for this very reason that any new systems we purchase are based on Windows Compute Cluster Server 2003.”
A case study about the Daggett Research Group’s work with HPC is available at http://www.microsoft.com/casestudies/casestudy.aspx?casestudyid=4000002104.
At the University of Washington and the University of North Carolina at Charlotte, professors are integrating HPC into their teaching. They are now using the Microsoft HPC++ Computational Finance (CompFin) Lab, an online HPC service aimed at providing academic computational finance programs with a rich analytic and simulation environment in the context of real market data. Professors are deploying their models via Microsoft Office SharePoint Server to a 256-core cluster, while students use Microsoft Office Excel to specify their experiments, seamlessly submit jobs to the cluster and visualize the results in their projects. Students at the University of North Carolina used the HPC cluster to examine mortgage pricing techniques as well as the effect that mortgage default and prepayment have on mortgage pools. More information about the CompFin Lab is available at http://labs.microsofthpc.net.
The Louisiana State University Center for Computation and Technology (CCT) has organized an HPC boot camp where high school students are exposed to the world of HPC and how it can be used to solve real-world problems. The CCT uses the CCS 2003 operating system running on Dell server computers to give students the opportunity to build their own compute nodes, connect the compute nodes together to form small HPC clusters, and run simple codes on the clusters. After the event, the HPC clusters assembled by the students are given to their respective high schools for continued use. A case study about the CCT is available at https://www.microsoft.com/casestudies/casestudy.aspx?casestudyid=4000001782.
At the University of Iowa, CCS 2003 is being used for HPC efforts in medical research and specifically for improving digital medical imaging. The increased tomography image resolution generated from HPC is helping to improve digitalized diagnostic systems for cancer and other health issues. More information about HPC efforts at the University of Iowa is available at http://www.uiowa.edu/~mihpclab.
“These institutions and others across the country are doing incredible work in areas that will change the world for future generations,” Salcito said. “Now, through one Microsoft platform, we’re enabling this progress — from deployment and parallel programming to management — in ways that will set their work apart in the supercomputing arena.”
To learn more about Microsoft’s HPC solutions for education, science and research, visit http://www.microsoft.com/hpc/education.mspx.
Founded in 1975, Microsoft (Nasdaq “MSFT”) is the worldwide leader in software, services and solutions that help people and businesses realize their full potential.
Note to editors: If you are interested in viewing additional information on Microsoft, please visit the Microsoft Web page at http://www.microsoft.com/presspass on Microsoft’s corporate information pages. Web links, telephone numbers and titles were correct at time of publication, but may since have changed. For additional assistance, journalists and analysts may contact Microsoft’s Rapid Response Team or other appropriate contacts listed at http://www.microsoft.com/presspass/contactpr.mspx.