Boise State University researchers generate large amounts of data, but often face obstacles when it comes to accessing and analyzing that data effectively. That’s about to change, the university says.
A group of Boise State University faculty from engineering, biological sciences, geosciences and computer science has received a grant from the National Science Foundation to build a high-performance computing and visualization instrument.
When the project is completed, researchers across multiple disciplines will have access to vastly improved capabilities for tackling large computational problems.
“The benefits of this grant will be felt far beyond Boise State University,” says Amy Moll, dean of the College of Engineering. “The plan is for the parallel computing and visualization cluster to be housed at a facility open to university researchers, as well as local technology companies and partners. This advanced cyber-infrastructure resource has the potential for a huge impact on our regional economy.”
The $555,384 grant was awarded under NSF’s Major Research Instrumentation Program. The money will be used to build a 32-node GPU/CPU cluster with a data storage array and a 5-foot-by-8-foot tiled display in a visualization theater setting.
GPU computing uses graphics processing units — GPUs — together with conventional central processing units — CPUs — for faster processing of computational science and engineering problems. The GPU/CPU cluster will support parallel computing and rendering, data storage and high-resolution imaging.
“Without supercomputing resources, computational analysis and massive data stores can become more of a burden than a help,” says Inanc Senocak, associate professor of mechanical and biomedical engineering and the principal investigator on the project. “This new computing cluster will extend our range of exploration in science and engineering projects and substantially accelerate the time it takes to get results.”
Senocak and co-investigators Peter Mullner in materials science and engineering, Hans-Peter Marshall in geosciences, Julie Oxford in biology and Tim Andersen in computer science have proposed using the computing cluster to support a range of research projects. They may include wind-energy forecasting, modeling for threat reduction in chemical and biological defense, materials characterization and modeling, snow hydrology and remote sensing, and mechanisms of skeleton development in living systems.
Senocak says the researchers also plan to make this cyber-infrastructure accessible to high-school science, technology, engineering and math scholars through outreach activities such as hands-on exercises for modeling and simulation, visualization of earth and space scientific data and high-resolution imagery.
“If we look beyond the obvious benefits to the researchers, we can only imagine the profound impact this kind of experience might have on the next generation of scientists and engineers,” he says.
A planned computing and visualization center will be open to local tech companies and high-school classes.