Pitt-Bradford students have been learning Aristotle’s adage that “the whole is greater than the sum of its parts” with a super computing project.
Working with Robert Ellison, systems architect and adjunct professor, students in the Computer Information Systemsand Technology program, created a small supercomputer capable of calculating pi 30 times faster than a desktop computer.
A super computer is a network of hundreds of computers performing part of a task simultaneously and then reporting back their results to the “head computer.” In this case, small, cheap computers called Raspberry Pis were used.
On the market for less than a year, Raspberry Pi is a $35 computer designed by a nonprofit to make technology available to younger, poorer students. The entire computer is about the size of a deck of cards and can be connected to a television screen for use as a monitor. An SD card like those used in digital cameras serves as its hard drive.
For college students, the affordability of Raspberry Pi provides more opportunities to network computers together, write computer code and practice solving real-life problems.
Ellison saw such an opportunity after reading an article about the University of Southampton using 64 Raspberry Pis networked together to create a supercomputer. He and Don Lewicki, associate professor of business management and director of the CIST program, decided a smaller version would benefit Pitt-Bradford students.
Senior Wes Milliron, a senior CIST major from Julian, jumped on board along with several other CIST majors: Jordan Reed of Genesee, Dan Comes of Cyclone, Tyler Morris of Bradford and James Rosenblatt of Wexford.
Building the computer took place outside of classes or clubs – just a bunch of computing lovers working on a project with Ellison there to lend a guiding hand.
Like any project, the devil was in the details. After purchasing 32 Raspberry Pis and making enough cable to wire them together, students had to figure out how to house and connect them so that they would communicate correctly with each other.
The next challenge was installing basic software. First it was loaded onto one computer. Students customized the software and then wrote programs to copy the customized software onto all 32 computers.
Milliron did much of the programming using a programming language he was teaching himself called Perl. Having a real project to work on made learning Perl much easier, he said.
After the machine was booted up for the first time, the team discovered more software was needed, and the process was repeated.
That kind of real-life problem solving is invaluable for students as they head out into the workforce.
“This is an eye-opening resume entry,” Lewicki said of working on the Raspberry Pi project.
Milliron agreed. By February of this year, he had secured a post-graduation job with the information technology consultant Nexstara in Mars. He said that his work on the supercomputer “was a big selling point” and that it set him apart from other candidates.
The computer should now continue to benefit students and faculty across disciplines who may be interested in using it for their own research.
Ellison has begun researching fractals, which are used to create models of things from the natural world such as mountains.
The computer can examine thousands of records for molecules, DNA, demographics or whatever a faculty member or student might dream up. That person could then work with students and faculty from the CIST program to write the program that would calculate the solution. Because the Raspberry Pis are so small, the whole computer is about the size of a filing cabinet and is portable.
Lewicki has an even bigger project in mind. With a computer lab scheduled to have its computers replaced this summer, he’s looking for the right student to create a super computer out of the 24 being replaced.
Milliron, however, is already busy.
For more information on the Raspberry Pi project or CIST program at Pitt-Bradford, contact Lewicki at (814)362-0988 or firstname.lastname@example.org.