IE 11 Not Supported

For optimal browsing, we recommend Chrome, Firefox or Safari browsers.

University of Idaho Competes with Big Leagues in Computing

Every processor can access all the memory, which makes it tremendously fast.

It might not be much to look at but a new computer at the University of Idaho (UI) can run simulations most computers across the nation cannot physically handle.

UI faculty have been experimenting with the supercomputer since the fall and have started to integrate its abilities into their research this semester. The computer, nicknamed "Big-STEM," provides researchers with an enhanced amount of memory to support large-scale simulations for the science, technology, engineering and mathematical divisions.

"It is a very rare type of machine, especially for a school like ours," said Jim Alves-Foss, director of the Center for Secure and Dependable Systems at the university. "It is primarily used for simulating and modeling."

The machine, which is about the size of an average computer tower, has eight 10-core processors and 4 terabytes of memory - and is expected to double by summer 2014.

The first part of the machine was funded from a $300,000 grant from the National Science Foundation. They have now received another $240,000 grant from the Murdock Charitable Trust and are hoping to match funds with the Idaho Global Entrepreneurial Mission to add another 4 terabytes of memory - giving the machine 4,000 times the memory of the average home computer.

"It is unique because every processor can access all of the memory, which makes it tremendously faster than anything we have had before," Alves-Foss said.

Mathematics professor Lyudmyla Barannyk has started using the machine to test the efficiency and accuracy of math algorithms. In the past, she has had access to systems at the Idaho National Laboratory but those computers did not have enough memory to support her research.

"It is more powerful than anything we know of," Barannyk said. "It opens up huge possibilities we couldn't even think of before."

Before the arrival of the machine, researchers had to simplify and downsize their models in order to run them on existing software. Often they had to back up their work to disks and run the models in different sequences. Because of the massive amount of memory storage, the Big-STEM allows the different simulations and models to be run simultaneously.

"We have some projects that would take weeks to run that are now down to hours," Alves-Foss said.

Several faculty members have already put the computer to work, which requires a constant cooling fan to keep the machine from overheating.

Alves-Foss said a physicist is currently working on modeling complex proteins and their interactions with chemicals in the human body to be used in the design of targeted drugs.

The mechanical engineering department is working on off-shore wind turbines on waves. The computer can simulate the complex motion of the waves and the wind modeling to help further understand building in that specific type of environment.

Researchers access the machine from their own computers, using the software and programs already installed on their devices to run by remote access on the supercomputer. They schedule use of the computer based on the amount of memory needed to run their project which allows multiple faculty members to use the machine at the same time.

Alves-Foss said aside from the supercomputing centers at select colleges around the nation he does not know of any other computer like the Big-STEM in the northwest and there are few others in the country. The UI does not currently plan to acquire more supercomputers but they have not ruled it out for the future. Alves-Foss said they want to monitor the Big-STEM before adding more machines.

Ray Anderson, information technology resource manager for research in the College of Engineering, called the machine "turn-key high performance computing" meaning they can take the computer right out of the box and begin using it because it does not require any specialized software.

"It allows researchers to start doing the research instead of spending time figuring out how to use the software," Anderson said. "This is the next step in the process - simplification."

©2014 the Moscow-Pullman Daily News (Moscow, Idaho)