Iowa State University

ITInformation Technology

Real-World Physical Phenomena: Model, Design and Control through HPC

This news item expired March 7, 2014. It may contain out-of-date information.

As High Performance Computing becomes more powerful, the programs and people working with such technology are seeing their passions become ever more relevant -- not only as independent research tools, but to the public as well.

Iowa State University's High Performance Computing (HPC) initiative includes a collection of powerful supercomputers that support the researchers and students who use them. And every few years, the university and its cohort of HPC users work with a combination of grant money and proposals in order to get a newer machine in its place. There are several courses offered at Iowa State that showcase how to use HPC.

Baskar Ganapathysubramanian, Assistant Professor in Mechanical Engineering at Iowa State, is an ISU HPC veteran.

Baskar Ganapathysubramanian, Assistant Professor in Mechanical Engineering at Iowa State

"HPC is a great enabler to computationally solving problems that cannot be conventionally solved," Ganapathysubramanian said.

Computers like CyEnce, a machine acquired by the university in 2013, are able to solve problems that no ordinary desktop or laptop can handle. This is accomplished by taking advantage of the processing power at hand, and creating enormous amounts of data for the researcher.

Ganapathysubramanian's group uses high-performance computing to model, design and control real-world physical phenomena. Using the power of the supercomputers such as CyEnce, his group's effort is directed towards understanding and improving clean energy applications.

For example, Ganapathysubramanian's group in collaboration with groups in Architecture (Ulrike Passe), Electrical Engineering (Umesh Vaidya), and Material Science (Krishna Rajan) design greener buildings (using clean energy technology) through HPC. HPC clusters can enable generation of extremely detailed information about variables such as airflow patterns, local temperatures, and comfort metrics for a fraction of the cost of actually measuring them. This allows exploring various designs and can help quantify potential energy savings. Another example is the group's use of HPC for high throughput computing to tailor processing conditions that will result in novel polymer morphologies. High performance computing platforms like CyEnce greatly facilitate accelerating the pace of discovery and deployment of novel, advanced materials. This is a key aspect of the Materials Genome Initiative.

Ganapathysubramanian says that most of the credit for the work should go to his students and post-doctoral associates. "They're a dedicated bunch of people who focus on the question, 'How can we use HPC to solve problems relevant to society?' Everything that we do is run by their enthusiasm for HPC," he said.