ENGINEERING

Duquesne University, under a National Science Foundation Grant (NSF), has initiated a unique learning opportunity for chemical engineering undergraduate, post graduate and doctoral fellows by purchasing two SGI Altix supercomputers and an SGI InfiniteStorage system from Silicon Graphics. Dr. Jeffry Madura, associate professor and chemistry chair at Duquesne, and his colleague Dr. Jeffrey D. Evanseck, associate professor of chemistry, have teamed up to create a state-of-the-art learning and research environment to enable their students to conduct research of the highest caliber in both quantum chemical computation and molecular dynamic simulation. A sampling of current projects the SGI Altix supercomputers are allowing students to delve into include: -- the structure and dynamics of DNA, with a focus on drug design for possible genetic control -- researching peptide encapsulated cadmium sulfide nanoclusters -- known as quantum dots -- which may form the basis of new optical devices -- a carbon dioxide sequestration project for the National Energy Technology Lab (to "sequester" carbon dioxide emissions away from the atmosphere in an effort to slow global warming) "We selected SGI Altix because, first of all, it was the fastest machine for the problems that we have. It wasn't theoretical speed: it was actual speed," said Dr. Madura. "There was extensive benchmarking and SGI Altix provided the fastest turnaround time; it was clearly the best solution for the kinds of jobs we want to solve. Whether it's quantum mechanical or classical mechanical, SGI Altix did the best for our applications. We also had the best working relationship with SGI. Plus, by partnering in their Innovators' Program we had the added confidence that SGI will make certain that this machine will run and perform the way they said it would. And, they've just gone beyond what I would think would be required. The machine is up and running and SGI was there every step of the way to make sure it running optimally. In every way, SGI and the Altix were the best choice." Duquesne University, through the MDA CAD Team at SGI reseller Clarix Technologies, Inc. in Rochester, NY, selected a SGI(R) Altix(R) 3000 supercomputer based on 32 Intel(R) Itanium(R) 2 processors, a 16-processor SGI(R) Altix(R) 350 and a SGI(R) InfiniteStorage TP9100 with their storage capacity currently over 2 terabytes. Duquesne runs different types of software on the Altix for two different types of research projects. If the research is quantum mechanical in nature, Gaussian software is employed. As an added bonus for students, Dr. Doug Fox, one of the founders of the Gaussian(R) software package, is part of the faculty at Duquesne. If the project is classical mechanical, researchers and students primarily use CHARMM (Chemistry at HARvard Molecular Mechanics) or AMBER (Assisted Model Building with Energy Refinement). The chemical computing group also utilizes a package known as MOE (Molecular Operating Environment). The Altix systems and storage work seamlessly with other lab systems on the network and visualization is achieved on the 22 PC-based laptops in the lab. "We're enhancing the experimental side of chemistry in a novel way, by teaching students how to use computers to be better experimentalists. It's an excellent fit with the way we carry out education, a synergistic relationship between research and education. Not only do we carry out nationally important research but we're training the scientists of the future," said Dr. Evanseck. "We're giving them a new skill that they may not have had access to learn before, so that they can become better chemists. We carefully craft out possible research projects for them so that they have to use the power of the SGI Altix supercomputers to make progress on a research project. We hold a weeklong workshop to give them the skills necessary and then we assign graduate student to assist them, plus experimental mentors who have experience to help them be successful, and Jeff Madura and I serve as theoretical mentors. It's a 3 on 1 situation that's greatly enhanced by SGI Altix because it enables us to do more projects more quickly, and because of that, we expect to obtain other grants to fund additional projects and research." Thanks to the NSF grant as well as another grant from the Department of Education, 15 students per summer are involved in this intense learning experience. In addition to multiple projects in chemical science, the department also collaborates with faculty from math and computer science, biological sciences, physics, and the school of pharmacy, using the SGI Altix supercomputers to solve real problems in chemistry, pharmacy, biology, and physics. "SGI is pleased to be part of this unique environment at Duquesne University, and we're equally pleased to welcome them to the SGI Innovator Program, which establishes long-term research and education relationships between SGI and academic institutions, national labs, government research facilities, and commercial science and engineering research centers," said Thomas Stanley, Director, Science and Technology, SGI. "The affordability of the SGI Altix 3000 family of supercomputers and their unparalleled performance is capturing the imaginations of the leading researchers, scientists, engineers and artists around the world, delivering mission critical support to defense and security, science and research, manufacturing, energy and media industries. We at SGI are very excited that the Altix systems are fueling student and faculty projects at Duquesne that will lead to greater understanding of our planet and humankind as well as scientific breakthroughs that could substantially enhance life on earth."