Purdue Computing Research Institute
Scientists and engineers rely on high-performance computing to conduct their work, yet every hour they spend refining their computing capability is an hour lost to their research. It's a problem in the sciences that is being addressed by the Purdue Computing Research Institute (CRI).
Thomas Downar, professor of nuclear engineering and codirector of the institute, says that the purpose of CRI is to conduct interdisciplinary research in computational science and engineering.
"CRI brings together researchers and builders of the enabling technology (advanced computer architecture, parallel numerical algorithms, system software, visualization, and data mining) with researchers in the various science and engineering applications," Downar says. "Our goal is to address the entire computational process—from the inception of the mathematical model to data analysis and visualization."
 Christoph Hoffmann, a professor of computer science who codirects the Institute, says that CRI works closely with the staff at ITaP to explore new ways for scientists and engineers to effectively harness computational power. "It's an important interaction," he says. "ITaP provides the needed high-end computational infrastructure to Purdue's science and engineering researchers, and by working together, CRI and ITaP enable these researchers to effectively utilize these advanced computing platforms."
In 2003, CRI and ITaP established a joint research position and brought Faisal Saied to Purdue as a senior research scientist. Saied built the nationally-recognized performance engineering and computational methods group at the University of Illinois ' National Center for Supercomputing Applications (NCSA).
Several collaborative projects already are underway, including research in computational nanoelectronics, structural mechanics involving LS-Dyna, and developing parallel linear systems and eigenvalue problem solvers for implementation on various high-end computing platforms.
"Through CRI, and in collaboration with ITaP, we are building an interdisciplinary simulation and modeling research community at Purdue," Hoffmann says. "This community will benefit interdisciplinary investigations in many fields of science and engineering." |
The more you know, the faster you go
Taking a trip means plotting a map and determining how to get from here to there quickly. Detours and traffic congestion are likely. Knowing about these things ahead of time allows for changes to make the trip easier. A similar process occurs with technology research and product marketability.
A technique used by business and research organizations for planning, technology roadmapping links together in one database areas such as research, product and market strategies, suppliers, and customers. Roadmapping is storing information in a common format and allowing easy access. The Center for Technology Roadmapping (CTR) advances this concept.
Edward Coyle, professor of electrical and computer engineering and assistant vice president for research computing; and Jonah Duckles, coordinator, oversee the operations. Duckles has a joint appointment between CTR and ITaP.
"Roadmaps enable companies to communicate visions, attract resources, stimulate investigations, and monitor progress," says Duckles. "Robert Galvin, former CEO of Motorola, made roadmapping popular. We are benefiting from Motorola's experiences."
CTR, Motorola Inc., and The Learning Trust (makers of the roadmapping software Vision Strategist) have partnered in North America for technology roadmapping capabilities. Efforts are under way to have similar partnerships in India, Australia, the United Kingdom, the European Union, and China.
Purdue also is partnering with Tsinghua University in Beijing to create a World-Wide Technology Roadmap Archive."This is important because perspectives differ around the globe as to what factors relate to each other in a roadmap," Coyle says. "Having a global view opens up a world of possibilities." — Jeff Buchanan |
