The supercomputer comprises of 64 processors, 1TB of memory (16GB SD Cards in each of the Raspberry Pi) and can be powered on using just a single 13 Amp mains socket. Message Passing Interface is used for communications between the nodes through the Ethernet port. The team managed to build the core of the supercomputer under £2500 (excluding switches for networking).
Other than Cox the team had Richard Boardman, Andy Everett, Steven Johnston, Gereon Kaiping, Neil O’Brien, Mark Scott and Oz Parchment as its members. The rack for the supercomputer has been built using Lego under the guidance of Professor Cox’s son James Cox (aged 6). In a press release Professor Cox said, “As soon as we were able to source sufficient Raspberry Pi computers we wanted to see if it was possible to link them together into a supercomputer.”
Named “Iridis-Pi” after University of Southampton’s supercomputer Iridis, software for the supercomputer has been built using Python and Scratch. Professor Cox used the free plug-in ‘Python Tools for Visual Studio’ to develop code for the Raspberry Pi.
“We installed and built all of the necessary software on the Pi starting from a standard Debian Wheezy system image and we have published a guide so you can build your own supercomputer”, Cox added.
Professor Cox adds, “The first test we ran – well obviously we calculated Pi on the Raspberry Pi using MPI, which is a well-known first test for any new supercomputer.”
The team wants to see “Iridis-Pi” become an inspiration for students to enable them “to apply high-performance computing and data handling techniques to tackle complex engineering and scientific challenges.”
[Update@18:43 UT]: We have been able to get in touch with Professor Cox and we are in the process of updating the article with his inputs on speed and performance of the supercomputer. Please, stay tuned.
[Update@18:52 UT]: According to Professor Cox, as of now the team hasn’t gone ahead with “any systematic testing on the whole 64 processor system”. Professor Cox is of the opinion that “the processor is probably not the place to get performance from on the Pi”. Iterating the point made earlier that the team was looking to harness the capabilities of such a marginally priced computer and wants the supercomputer become an inspiration for students, the professor said, “we were interested in how the price to get a system together has gone from 1,000,000s to 100,000s, to 10,000s and now down to 1000s of dollars/ pounds and that with Raspberry Pi you can really show (non virtualized) parallel programming in the raw at the 1000s dollar price point.”
In regards to system images for Raspberry Pi, the professor said, “…the core operating system images that are available are still being nicely tuned up with some boosts in performance between releases of the system images along with other kernel/ firmware updates and enhancements.” Cox said that the switching fabric used to connect 64 Raspberry Pis also plays an important role when it comes to determining performance of the supercomputer. The professor ended the email with “There is certainly quite a lot of interesting work to be done on performance – and something that anyone can get on with!”
Thanks Professor Cox.