NSF supports creation of science and engineering high-performance scientific computing cluster at UWM

Ryo Amano Mahsa Dabagh Philip Chang Clark Evans
Ryo Amano, Mahsa Dabagh, Philip Chang and Clark Evans are part of the NSF Campus Cyberinfrastructure Program at UWM

High-performance scientific computing clusters are helping to speed the process by which scientific researchers collect and analyze enormous amounts of scientific data. In large part because of the vast amount of data available, these clusters are increasingly seen as necessary for breakthrough discoveries and their creation at universities—including UWM—is being supported by the National Science Foundation’s Campus Cyberinfrastructure (CC*) Program.

Recently, two faculty members from the College of Engineering & Applied Science became part of an NSF grant to help create such a cluster at UWM, one that will aid faculty researchers—in fields including astronomy/astrophysics, atmospheric sciences, biomedical sciences, engineering, freshwater sciences and physics—and students taking computationally intensive classes.

Ryo Amano, professor, mechanical engineering, and Mahsa Dabagh, assistant professor, biomedical engineering, are co-PIs on the $400,000, two-year grant from the NSF’s CC* program.

Philip Chang, associate professor, physics, is the PI; Clark Evans, associate professor, atmospheric sciences, joins Amano and Dabagh as a co-PI.

The team’s project—CC* Compute: A Balanced Cluster for Science and Engineering in the Great Lakes Region— will support UWM as it builds and deploys a high-performance computing cluster that meets science researchers’ needs for computing, processing speed and memory.

UWM already operates a high-performance cluster (a group of computers that work together to solve problems). The new, high-performance computing cluster will be a group of computers designed to tackle scientific problems using computational methods.

About UWM’s science and engineering high-performance cluster

According to Amano, the cluster will help researchers push the computational frontiers of their research, create cutting-edge computational tools, and scale their computations to run on other national supercomputing centers including the NSF-funded Extreme Science and Engineering Discovery Environment (XSEDE).

It will be housed in a designated computing room in the Kenwood Interdisciplinary Research Complex and integrated into the Open Science Grid in order to enable full utilization of idle computing resources to enhance local, regional, and national scientific computational capabilities.

Faculty researchers and students will have access to high-performance computing 

In UWM’s College of Engineering & Applied Science, the new cluster, which will increase UWM’s computing capability by tenfold, is expected to benefit faculty researchers and undergraduate and graduate students.

For example, Amano relies on high-fidelity computations for his research on renewable energy sources and fluid mechanics. Such computing power is necessary, he says, to perform large-eddy complex turbulence simulations, multiphase flows, and chemically reacting flow computations.

“All the computations performed in the supercomputing clusters can be replaced by the new cluster system,” he says.

Dabagh, in her research on biomedical engineering, expects that the new computing cluster will speed along her research on computational models that predict the progression of specific diseases—such as cancer and atherosclerosis—in an individual.

“The new cluster will ease our efforts in creating large-scale studies of organs with complex anatomy, including small capillaries,” she said.

Students also will benefit from using and understanding high-performance computing, a technology that is establishing its place on the cutting edge of scientific research. “Most of my doctoral students work on computational studies to validate their sophisticated experimental results on aerospace, energy and water research,” Amano says. “Currently, the process takes days for a single run. With the new cluster, we can achieve such jobs in a day or even a few hours.”