COLLEGE PARK, Md. – Big data is a big deal these days whether scientists are studying black holes, the trillions of microorganisms that live within our bodies, or climate change impacts. Researchers who use large data sets to study these and many other topics increasingly need more powerful computers and more digital storage space. To address this demand, two of Maryland’s most powerful research universities are preparing to open one of the largest academic computing centers in the nation.
Supported by $30 million in state funding, the Maryland Advanced Research Computing Center (MARCC, pronounced “MAR-see”) is expected to provide state-of-the-art digital processing power to a wide array of researchers at the University of Maryland and Johns Hopkins University. The computing center is located in Baltimore near the edge of the Johns Hopkins Bayview Medical Center campus. Final testing at the facility is under way, and it is expected to be fully functional in early July.
The shared equipment within the new facility will be capable of delivering a hefty digital punch. The setup includes more than 19,000 processors and 17 petabytes of storage capacity—that’s 17 million gigabytes.
“In recent years, the rapid evolution of big analytics technology has vastly expanded the ability of researchers to discover new knowledge in almost every discipline - from the natural sciences to human behavior, and economics to education,” said University of Maryland’s Vice President for Research and Chief Research Officer Patrick O’Shea. “Taking advantage of the revolutionary potential of research involving large data sets to transform knowledge and improve human lives requires expanding the computing resources available to researchers. This new joint supercomputing center will do just that.”
Thanks to speedy fiber-optic cable connections to the participating campuses, researchers at JHU and UMD won’t have to leave their labs or offices to tap into the new computing center.
“Everyone is going to be able to access the new facility on a remote basis,” said Jaime Combariza, a Johns Hopkins computational chemist who became director of MARCC in June of last year. “MARCC allows all of Johns Hopkins and the University of Maryland to centralize their computing power.”
For participating researchers, the arrangement should lead to significant cost savings and greater efficiency, according to Combariza. Instead of requiring individual research groups to use time, money and space to create their own high-performance computing centers, all participants will share the costs of cooling, networking and running the single center.
The field of supercomputing is well known for engineering extreme processing speeds. However, increasingly, researchers’ calculations are limited not by the speed of processing, but by access to and efficient use of unparalleled amounts of data. The many UMD research areas where advanced data and computing technology is vitally important include: genomics, astronomy, neuroscience, climate sciences, cyber security, language science and health sciences.
The Maryland Advanced Research Computing Center, adds significantly to the existing supercomputing capabilities of the University of Maryland, including its largest supercomputer Deepthought2 launched in May of 2014. Deepthought2 can complete between 250 trillion and 300 trillion operations per second. It has a petabyte (1 million gigabytes) of storage and is connected by an InfiniBand network, a very high-speed internal network.
Among the UMD scientists who look forward to using the center’s new supercomputing power is Distinguished University Research Professor of Mathematics and Physics James Yorke, known worldwide for helping name and lead the development of “chaos” research – the mathematical study of complex (nonlinear) dynamic systems like weather and the spread of epidemics.
Yorke said one of the projects for which he hopes to use the new center is ongoing pine tree genome work he is doing with colleagues from the University of California, Davis and Johns Hopkins University. Last year their team assembled the genome of the loblolly pine, which, at seven times the size of a human's, was the largest genome ever sequenced. Now they are working on the sequencing of a type of pine tree with an even larger genome. “This is a big computational project and we are delighted to get the additional computational resources of this center,” Yorke said.
According to Yorke, such work both advances genomic science and has potential economic benefits because pines are a major source for timber and paper products. Understanding pine genetic codes “may be helpful in breeding trees with improved growth rates and resistance to disease,” he said.
Read more about UMD’s high-impact, big data research here:
July 7, 2015