MCLA’s ‘Supercomputer’ an Invaluable Research Tool
Dr. Elizabeth Hartung, an assistant professor of mathematics, traveled to Nankai University in Tianjin, China, this summer, where her research on “Pairwise Incompatibility,” conducted at MCLA using the campus’s supercomputer, was presented. In addition, she presented new information from her innovative research in mathematical chemistry and graph theory.
At the Annual Meeting of the International Academy of Mathematical Chemistry, and the 2016 International Conference on Mathematical Chemistry, Hartung offered a presentation on “Internal Kekule Structures for Graphene and General Patches,” a result of joint work with mathematics professor Dr. Jack Graver of Syracuse University in Syracuse, N.Y.
Hartung uses mathematical models of carbon molecules – such as graphene and nanotubes – to understand properties related to the stability and conductivity of the molecule.
The research problems she and Graver studied came from questions raised by a chemist and a mathematician at a previous conference. One researcher who attended this recent presentation commented that, before going to her talk, he did not believe the results could be true. However, after he heard what Hartung had to say, he realized that her work was “something the field had missed.”
Discovering the new findings, “was exciting because it was unexpected,” Hartung said.
A second presentation centered on research conducted on MCLA’s supercomputer, which the campus acquired in 2010. Since its launch, it’s been used in projects by members of MCLA’s mathematics, biology, physics and chemistry departments.
“Pairwise Incompatibilty of the Kekule, Fries, and Clar Numbers for Benzenoids,” was the result Hartung’s work with MCLA alumni James Chapman ’15 and Andrew Nelson ’15, back when they were students. Additional research was conducted by computer science faculty member Dr. Aaron Williams and student Judith Foos from Bard College at Simon’s Rock in Great Barrington, Mass.
Hartung explained that Chapman and Nelson wrote a computer program and found the first counterexamples. Then, Williams and Foos created a more comprehensive computer program to find all counterexamples up to a certain size of molecule. All of them used MCLA’s supercomputer for the computations.
As a result, “We found examples of pairs of graphene patches where these parameters were ‘incompatible,’” Hartung said.
MCLA’s supercomputer, explained Ian Bergeron, director of information technology, is a high performance computer cluster (HPC). The cluster of systems shares computing power and data storage over a high speed data network.
“This is a very useful architecture for an educational institution, because it allows users to scale their projects up and down by more or fewer of the computing nodes. It also lends itself well to many simultaneous users,” Bergeron said.
“We share access to the system with Nuclea, but it resides in our datacenter where our faculty and students can access it readily,” Bergeron added.
“Data analytics has become deeply embedded in many disciplines and this system is an invaluable tool for our faculty and students to incorporate ‘big data’ and parallel computing in their own research.”