CULLOWHEE - A cluster computer developed in Western Carolina University’s department of mathematics and computer science was part of a worldwide demonstration of grid computing at SC 2003, the premier conference on supercomputing, held recently in Phoenix.

The cluster computer at Western was connected to other computers from 21 countries to form a computational grid used to solve problems in high-energy physics, molecular docking, portfolio analysis, neuroscience and natural language processing. The cluster computer is called leo.cs.wcu.edu-or “leo” for short.

“Leo” is part of a grid computing laboratory established by Western faculty members Barry Wilkinson, Mark Holliday and David Luginbuhl, associate professors of mathematics and computer science. A team of mathematics and computer science students-Siarhei Liakh, Shane Bradley, and Bruce Hauman-and Kyle Marks, a student from Smoky Mountain High School, has led the implementation of the leo cluster.

“Grid computing is widely viewed as the next step in the evolution of the Internet, as well as a natural development from cluster computing,” said Wilkinson. “In cluster computing, computers at a single location are grouped together to form a high-performance computing platform. With grid computing, the cluster becomes a global unit with computers distributed worldwide.”

The word “grid” was borrowed from the electrical industry, he said. The power grid distributes electricity, as generators on the grid are used collectively to supply the demand for electrical power. Similarly, computers in grids are used collectively to satisfy the demand for computing power, Wilkinson said.

Grid computing offers the potential for more efficient use of computing resources both within organizations and between organizations, as clusters of less-expensive computers are linked worldwide to provide computing power at a level previously possible only by using expensive supercomputers, he said.

Among the anticipated benefits of grid computing:

• Solutions to scientific, medical and engineering problems too complex to be solved by individual computers. For example, a grid is currently being used to employ computational chemistry to develop a medical therapy to combat the smallpox virus.

• Smaller institutions, by using a grid, would have access to expertise and computing resources impossible for each institution to accomplish separately.

• The core business processes of an organization can be re-implemented using a grid, resulting in dramatic cost savings.

Grid computing, the subject of a November gathering in Asheville, also could provide a major boost to the North Carolina economy. North Carolina’s Rural Internet Access Authority predicts businesses will become increasingly productive and profitable by taking advantage of grid computing resources. In North Carolina alone, forecasters estimate an additional $10 billion in economic output, 24,000 new jobs and $7.2 billion in higher wages by 2010.

Western is part of a newly launched University of North Carolina initiative aimed at developing a state computing grid, an effort that has received a $7 million pledge from MCNC, a nonprofit agency formed to advance technology-led economic development and job creation throughout North Carolina.

The UNC Office of the President convened a meeting in October of representatives of the 16 system campuses to discuss the potential grid computing offers to the state of North Carolina.

Western’s grid computing efforts also come as the university prepares to plug into a federally funded fiber-optic cable network that will provide the mountains of Western North Carolina with affordable, high-speed Internet access.

“Western’s grid computing laboratory will help Western North Carolina take the lead in the transformation of information technology through grid computing,” said Wilkinson.

For more information about grid computing, contact Mark Holliday in the department of mathematics and computer science at (828) 227-3951.

Maintained by the WCU Office of Public Relations
Last modified: Thursday, March 27, 2014 | Originally published: Monday, January 5, 2004
Copyright 2003 by Western Carolina University