A Leader in Supercomputing

A grant of $486,256 from the National Science Foundation will fund the purchase of a new high-performance computer cluster, making Oberlin a leader in supercomputing among liberal arts colleges.

The new computational cluster will be far more powerful than the existing 64-node “Beowulf” cluster acquired in 2005 with NSF support. At the time, Oberlin was among the first liberal arts colleges in the country to provide undergraduates with access to a supercomputer that can process gigantic data sets. Today, an increasing number of classes throughout all the scientific disciplines, including social sciences, have become reliant on supercomputing.

The cluster will be used by more than a dozen faculty and their research students to support an array of research projects ranging from modeling black hole collisions to analyzing the effects of organic molecules on air quality, and from determining the protein characteristics of the earliest life on Earth to reconstructing the evolution of flowering plants using massive amounts of DNA sequence data.

In addition to dramatically increased processing speeds, the cluster will provide at least 200 terabytes of long-term storage space.

Associate Professor of Biology Michael Moore, principal investigator for the grant, explains that two recent trends throughout the sciences—“big data” and the infusion of modeling—have led to an era where supercomputing is the norm.

“It's really imperative that Oberlin students get exposed to the world of high-performance computing,” Moore says. “Having a supercomputer on campus, instead of using off-campus computational clusters, will give students much greater access to high-performance computing than they would otherwise. There are lots of research-based supercomputers that are publicly available, but they aren't designed for dozens of students to log on simultaneously to run computational jobs, which is part of our aim for the new Oberlin cluster.”

Improvements in scientific instruments used in various disciplines have made it possible to generate enormous amounts of raw data that simply cannot be analyzed without the help of high-performance computers. “A great example of this is the genomics revolution in the life sciences over the past few years,” Moore says. “We can now sequence whole genomes routinely in a matter of days, but the analysis of the resulting huge data sets requires supercomputing. Likewise, intensive computational models are now standardly employed in fields as diverse as economics, atmospheric chemistry, and neuroscience, all of which are active fields of study at Oberlin.”

The cluster will be critical for both of Moore’s NSF-funded research projects. Two types of analysis employed in his lab—phylogenetic analysis and genomic/transcriptomic sequencing—increasingly depend on high-performance computing.

“The new cluster will enable sophisticated analyses of literally thousands of genes simultaneously. Just as importantly, the greatly expanded storage space of the new cluster will provide a safe, backed-up storage location for the huge data sets that are generated by the genomic sequencing projects in my lab.”

The grant proposal was co-authored by faculty members Matthew Elrod, Biggs Professor of Natural Science; Aaron Goldman, assistant professor of biology; Manish Mehta, Donald R. Longman Professor of Chemistry; and Robert Owen, assistant professor of physics and astronomy.