ATLANTA—Georgia State University has introduced the Advanced Research Computing Technology & Innovation Core (ARCTIC) to support research that would not be possible with standard consumer-grade computing, including analysis, modeling, simulation and the prediction of complex phenomena.
Now a resource for investigators at Georgia State, ARCTIC will soon be made available to scientists around the world.
ARCTIC was developed using a $1.2 million grant from the National Science Foundation and offers researchers advanced cyber infrastructure along with training and support. The system includes high-performance computing clusters and data storage systems, research networking and cloud computing.
“We don’t just provide the hardware,” said Suranga Edirisinghe, associate director of ARCTIC. “We work hand in hand with investigators and guide them how to best use the resources.”
ARCTIC is particularly aimed at investigators who are not traditionally served by high-performance computing, such as psychologists, biologists, neuroscientists or public policy researchers. The team also builds scientific gateways — easily accessible Web portals — to allow the public to access a project’s findings.
“These days, researchers have more and more data, but they may not have the technology to process that data. We’re trying to fill the gap,” said Edirisinghe. “The goal is to reach out to people who don’t traditionally use high-performing computing and bring them into the community, so they’re not constrained by resources. That’s also why the support piece is so important.”
“Anyone who has said, ‘We can do this on my office computer with a small sample, but to really be representative, nail down the model and generalize what we’re finding, we need to be running this on millions of data points,’ that’s who this is for,” said Jessica Turner, professor of psychology and neuroscience and lead investigator on the grant.
Turner is among more than 300 Georgia State faculty who have begun using ARCTIC to conduct their research. She is testing whether an algorithm can predict depression in various racial groups by analyzing brain scans. Faculty can also use the resource in the classroom to teach students about data mining or big data analysis approaches.