Computing cluster allows for interdisciplinary research
As a result of a previous National Science Foundation grant, the computer science department recently purchased a 12 node computing cluster to be used for computational research in mathematics, physics, mechanical engineering, computer science, and biology.
The computing cluster is essentially many processors clustered together in a small rack. Being more space and energy efficient than multiple computers, it is much easier to manage.
“This cluster has the equivalent computing power of 24 of the fastest computers on campus,” explains Chun Wai Liew, associate professor and head of computer science. “It is used to solve problems that require a lot of computer power, particularly if the problems are amenable to a parallel approach where multiple computers are used – each computer solves a small part of the problem and the pieces can be combined to form a solution to the larger problem. There are many problems like that, computer modeling and simulation of mathematical models being one of the common ones.”
Liew and Robert Root, associate professor of mathematics, received a NSF grant in 2005 to purchase multiple computers to help with the simulation of computer/mathematical models of swimming fish. The college aided in the purchase through committed funds to purchase new computers every two years, thereby keeping the resource up to date and state of the art.
The computer science department has made the cluster available to faculty and students in other departments to use for research.
“There are many problems that cannot be tackled/solved without significant computing resources,” says Liew. “This resource supports the faculty in their research on these types of problems. The computers can also be used to develop computational modules for different academic disciplines that can be used in the classroom.”
Liew says other faculty members who expressed interest in working with the 12 node computing cluster include: Tony Novaco, Metzgar Professor of Physics; Jenn Rossman, assistant professor of mechanical engineering; Steven Nesbit, associate professor of mechanical engineering; Derek Smith, associate professor of mathematics; and Louis Zulli, associate professor of mathematics.
Some of the ways in which faculty are thinking of using the cluster are computational art, modeling blood flow, and modeling swimming fish.
Liew hopes that through using the cluster he will be able to model the biomechanics involved in how fish swim, allowing for further investigation regarding the origin of vertebrae and fish locomotion. Since some of Liew’s models require 40 days to complete calculations, the computing cluster will allow data to be collected more quickly since 48 experiments can be run simultaneously.
Liew is also working with Jim Toia, director of the Grossman Gallery and Community-Based Teaching Program, to develop a computational model that emulates research completed on spore dispersal in mushrooms.