Young computer developers are always motivated to invent new technologies. The folks involved in the Parallella project managed to build an extremely powerful computer that has four more attributes: it’s small, it’s energy efficient, it’s cheap and can be used in parallel computing.
These days, building a supercomputer from pieces available on the market is a trivial operation that can be performed by anyone. Adapteva engineers took things to the next level, creating a small supercomputer that’s really affordable. The company is specialized in the manufacture of scalable multicore processor computer chips. They have created a computer the size of a credit card that can successfully perform tasks reserved normally to supercomputers when it’s connected with other similar computers. The company managed to raise just enough money for the project with the help of fans and gadget enthusiasts found on Kickstarter platform. Supporters of this project donated $900,000 via KickStarter in order to be able to get their hands on this small supercomputer. Adapteva delivered the first Parallella computers this summer and the batch for January 2014 is already sold out on preorders.
Another interesting fact about this project is the price. Although Adapteva has production costs of $150 per unit, they released the gadget on the market at the incredible price of 99 dollars. Company founder Andreas Olofsson told Ars that they will probably loose money in the first phase of the release, because the materials used to build Parallella are pretty expensive.
Guided by the tag line “a supercomputer for everyone”, each such Parallella comptuer features a E64G401 64-core microprocessor/coprocessor based on Epiphany™ multicore architecture clocked at a frequency of 800 MHz, 1.6 TB/s memory bandwith, 102 GB/S Network Bandwith, 2 Watt maximum chip power consumption, C/C++ programmable, chip-to-chip links for multimple connection.
The computer that’s the size of a card is not a supercomputer by itself , but rather a piece of a supercomputer. It can be used in applications involving smartphone, computing infrastructure, Mil/Aero, medical, communication, industrial and security projects.