University of Florida receives $500,000 ‘Big Data’ boost from IBM

September 25, 2014

GAINESVILLE, Fla. --- Imagine a day when your toothbrush can record every brushstroke you make and transmit that data to your dentist, who can use it at your next checkup to help you improve your brushing technique to prevent the onset of tooth decay.

That scenario is in the realm of possibility for researchers who work in the field of “big data,” a term for the collection of data sets so large and complex that they become difficult to process and analyze using traditional data processing applications.

Similar scenarios are one step closer to reality for University of Florida researchers, thanks to a partnership with IBM and its donation of a system that will enhance support for research and education on high-performance data analytics, or HPDA.

IBM’s PureData for Analytics system, valued at $500,000, will speed up processing capabilities and allow scientists working in many fields, including biological and environmental research, business infomatics and healthcare, to crunch data faster, said Elias Eldayrie, vice president and chief information officer of UF Information Technology.

“This is a great day for the University of Florida,” Eldayrie said. “The processing and analysis of our faculty’s research has been greatly increased, meaning their important work in health care, climatology, and other critical areas makes it into the global conversation faster. We have faculty conducting brain injury research. We have faculty studying how oil spills impact the fisheries along the Gulf of Mexico, and so many more projects of significant state and worldwide impact. Their research in these fields of study can get to the marketplace quicker thanks to the computing abilities we now have in place at UF.”

The PureData Analytics system, which will be managed by UF Research Computing, reduces the time needed to process large data queries of 10 million to 100 million records from 27 hours to a mere 3 seconds, a boon for UF researchers and the scientific community, Eldayrie said. The ability to process data faster will allow UF faculty to analyze their research more efficiently and share findings with top institutions, as well as the scientific organizations and government agencies sponsoring UF’s expansive research portfolio.

With Internet traffic estimated to quadruple by 2015, when nearly 3 billion people will be online pushing the data created and shared to nearly 8 zettabytes, the type of speed afforded by the PureData for Analytics system means there is a broader application of what can be done with high-performance computing, according to Sri Srinivasan, vice president and general manager for IBM’s big data platform.

“There’s definitely a big tsunami of data hitting us every day,” Srinivasan said. “When you take this paradigm and apply it to healthcare, retail or transportation, there are amazing opportunities to uncover.”

The PureData for Analytics system will complement UF’s HiPerGator, the most powerful supercomputer in Florida, which went online in May 2013. HiPerGator, along with a recent tenfold increase in the size of the university’s data pipeline through Internet2 connectivity, has already positioned UF as one of the nation’s leading public universities in research computing.

The addition of the PureData for Analytics system and its ability to optimize hundreds of millions of records will further UF’s goal of becoming a Top 10 public university, Eldayrie said.

“Over the past few years, UF has seen a 275 percent growth in research faculty using HiPerGator,” he said. “The addition of PureData for Analytics from IBM will add to that growth, enhancing the University of Florida’s ability to change the lives of millions of people.”

The new system, unveiled Tuesday at the Fall 2014 Research Computing Day, has been installed at the UF Data Center located on the East Campus.