More Big-Data Computing Benefits BusinessPosted on: April 4, 2012
Fossil fuel production, fishery viability, climate change and gaming development may seem like far-flung industries.
But they and other fields benefit from new computing power available at the University of Arizona’s University Information Technology Services (UITS).
A recent major expansion of the UA’s Research Data Center allows researchers to compile massive amounts of information in a matter of minutes or hours instead of days or weeks.
Only faculty-led research groups and students can use the center. Their experiences and findings can affect many types of business and industry.
For instance, UA geosciences professor Joellen Russell compiles data on climate change in various geographic settings. With super-computing power, she’s able to create models of climate stretching from the present back into the past and forward into the future.
Geologists with ExxonMobil use her findings and expertise on the historical interaction of mountains and climate to understand the climatic settings that are good for oil and natural gas formation.
Ecologists tap into her findings to project how habitats for penguins and whales may be altered because of climate change.
UA doctorate candidate Bonnie Hurwitz used the Research Data Center to compare 40 million known viral protein sequences to the 1.8 million she’s found on the ocean floor.
Understanding how the newly discovered protein sequences function could lead to ways of protecting commercial fisheries from deadly pathogens.
Creating Big-Data Workforce
Big-data understanding goes far beyond research universities and institutions. Company executives are realizing that a lot of information is available for interpretation, if only it could be compiled and analyzed.
Russell, who chairs the UA Steering Committee for Research Computing, knows of corporations building super-computing facilities to gather and analyze data in their respective fields.
“We can generate data like never before,” says Dr. Leslie Tolbert, UA senior vice president for research. “How do you turn it into useful information? We need to educate people who know how to do this in a real, serious way.”
The new Research Data Center ramps up the training of students in how to mine and analyze information buried in large, complex data sets.
“We had two students hired away by BP and ExxonMobil,” Russell says of a recent example. “They’ve been trained on the state of the art.”
She foresees demand for super-computing experts in retail, gaming, graphics and all fields of science.
Integrated Research Computing Services
The UA updates its research computing capabilities every three or four years. This year, the Research Data Center moved from cramped, shared quarters into a spacious, specifically configured room for computer processors, data storage and networking technology.
It currently has 4,324 processors. In comparison, a laptop has one, maybe two, says Dr. Michael Bruck, UITS assistant director.
“We can run large numbers of individual jobs or large parallel jobs,” Bruck explains.
And there’s room to grow. Researchers with publicly- and privately-funded grant projects can install computers they need for their respective projects.
They then get first access to the buy-in equipment. When those computers are not used by the project, they become available to everyone else.
Centralizing research computers not only provides more capacity, but allows researchers to pool administrative, maintenance and consulting talents and costs.
UITS provides other services that together span a research project’s major components.
Consultants help researchers create a statistic-gathering and -analysis plan and learn how to optimize use of the computers.
AZ-LIVE creates three-dimensional interpretations of data that then are projected in a way that immerses a person in the image.
For instance, contractors for the Arizona Stadium’s North End Zone project were able to physically, virtually “walk through” a computerized model of the architectural plans. Think “Star Trek” holodeck using glasses.
“When you’re close up in a 3D structure, you can see fine details you wouldn’t otherwise,” says Bruck.
To find out how the private sector can participate in research that requires super-computing, contact the UA Office of Corporate and Business Relations.