Lego. Суперкомпьютер на базе Raspberry Pi и Lego

[bga68]

11 сентября 2012 опубликован пресс релиз интересного проекта по созданию суперкомпьютера из 64 Raspberry Pi с подробнейшей инструкцией от создателя, профессора университета Саутгемптона Саймона Кокса — по этой ссылке...

В качестве элементов для создания «серверных шкафов» послужил обычный конструктор Lego.

В качестве ПЗУ для этой системы использовались карты памяти формата SD, объемом в 16 ГБ. Такие карты были установлены в каждую «Raspberry».

Как сделать такой компьютер?

Watch on YouTube

О компактном демонстрационном кластере на базе Raspberry Pi описано в статье: "Iridis-pi: a low-cost, demonstration cluster"

Авторы статьи:

Simon Cox is Professor of Computational Methods and Head of the Computational Engineering and Design Research group in the Faculty of Engineering and Environment at the University of Southampton. His research interests are in the application and development of computational tools, technologies and platforms to enable science and engineering research.

James Cox (aged 6) is at West Hill Park School, Titchfield, UK. He worked on this project at the University of Southampton over summer 2012 and is interested in computing and gaming using the Raspberry Pi.

Neil O’Brien holds an MPhys (Hons) degree from the University of Southampton and has recently completed his PhD in Algorithms for Scientific Computing at Southampton’s Faculty of Engineering and the Environment. His research interests include numerical methods, computational technologies, and smart home networks.

Mark Scott joined the University of Southampton in 2001 as an I.T. Systems Architect and Developer. After working for several groups in the School of Engineering Sciences, he moved to the central I.T. department and is currently working on the development of data management projects for the University such as the Materials Data Centre. He is also studying for a PhD at the Microsoft Institute for High Performance Computing in the Faculty of Engineering and the Environment.

Richard Boardman completed a PhD in Computational Physics in 2005. Since then, he has worked in computational chemistry and computer science, and currently works in X-ray computed tomography.

Steven Johnston completed an MEng in Software engineering and a PhD in data management. He is currently a senior research fellow at the University of Southampton and a program manager for the .NET Gadgeteer team at Microsoft Research. His research interests include embedded hardware, data management, the internet of things and cloud computing.