posted on February 22, 2013 |
| 7067 views
Despite its unusual name, the Raspberry Pi is one of the most popular computers in the world. With capabilities equivalent to a desktop PC, this $35, credit-card size machine is the favorite computing platforms for many hobbyists around the world.
At that price, it’s no wonder the Raspberry Pi boards have attracted interest from academia to teach computers and programming to novices.
Professor Simon Cox from the University of Southampton came up with a new twist for teaching with Raspberry Pi. He wanted to see if it was possible to link a number of boards together to create a supercomputer that would help students learn about parallel computing. If successful, it could even apply high performance computing and data handling to tackle complex engineering and scientific challenges.
After connecting 63 Raspberry Pi boards using Ethernet ports, he built a network of mini computers on a rack made out of LEGO blocks. His team installed and built the software based on a standard Debian Wheezy system image.
The LEGO racking of the supercomputer was a design developed by Simon Cox and his 6-year-old son James Cox, who has also been testing the Raspberry Pi by programming it using free computer programming software Python and Scratch.
The University of Southampton’s Raspberry Pi-based Supercomputer, named the “Iridis-Pi”, runs off a single 13 Amp mains socket and uses a technique called the Message Passing Interface (MPI) to communicate between 64 Raspberry Pi boards using Ethernet. The whole system cost under £2,500 ($4,200) and has a total of 64 processors and 1Tb of memory (16Gb SD cards for each Raspberry Pi). Professor Cox uses the free plug-in ‘Python Tools for Visual Studio’ to develop code for the Raspberry Pi.
The system is now capable of performing complex parallel computing tasks via splitting them amongst multiple Raspberry Pi processors.
Professor Cox has published a guide explaining his supercomputer and how he made it. You can learn and build your own supercomputer through this link: http://www.southampton.ac.uk/~sjc/raspberrypi/