Volume 2, Number 19 -- September 25, 2000. | ||
This past summer, Carleton University's School of Computer Science received more than $2.2 million from the Canada Foundation for Innovation (CFI) as part of a four-university consortium to establish the High Performance Computing Virtual Laboratory (HPCVL). The four universities are the Royal Military College, the University of Ottawa, Carleton University and Queen's University, the lead institution on the project. State-of-the-art multiprocessor computers on each campus will be linked by extremely fast and efficient communications switches, creating a network that will enable dozens of researchers to process massive amounts of data in a very short time.
"This gives a tremendous boost to the R&D activities carried out at Carleton University that are computationally intensive," says computer scientist Jörg-Rüdiger Sack, the NSERC Chair in Applied Parallel Computing. "The Laboratory can be used by Carleton researchers who need high performance computing power, and we invite industrial and government partners to make use of these facilities as well." The system allows Carleton's computers to access the machines located at the other three locations thereby greatly enhancing the computional power. Large modelling tasks, for example, can now be carried out in minutes as opposed to hours. Tasks that were previously infeasible can now be successfully attacked. Carleton will also be able to attract high qualified and talented researchers working in those areas. Sack also points out in terms of hiring new faculty the HPCVL will attract the very best. "It is a fantastic recruitment tool for new computer science and computational sciences faculty. They will be primary beneficiaries of this investment." At Carleton, the system is composed of a SUN SMP computer, a beowolf cluster providing 128 processors, and a large one terabyte server. As well, the funding will also support renovations of laboratory space. HPCVL will support innovative research across a wide spectrum of disciplines. Examples include the creation of new drugs, testing models of fluid dynamics to help in new aircraft design, the design of new polymer materials on the molecular scale, and mathematical models of the brain and human memory. "Parallel computing
has come a long way since its early days. Highly specialized
supercomputers have been replaced by general purpose computer technologies
incorporating fast switches," says Sack. "Clusters of workstations can
often provide inexpensive and quite powerful solutions to many industrial
and research challenges." More information is available at http://www.scs.carleton.ca/~sack.
|