LCE-EU 2015

In this talk we introduce an open source, home-brew, Beowulf architecture which was recently developed at the Bulgarian Academy of Sciences. The nodes of this minicluster are constituted of Radxa Pro single board computer, based on the (quadcore) ARM Cortex-A9 CPU. We demonstrate that it is possible to achieve performances exploitable even for scientific (computationally demanding) codes by running several parallel GNU packages such as nano-archimedes. In particular we show that it is possible to achieve very advanced simulations in the field of quantum computing.

Jean Michel Sellier will show how to use the previously described novel parallel hardware to simulate time-dependent quantum systems, which are well known to be computationally very demanding, such as candidate semiconductor devices for quantum computing (by exploiting single dopants at room temperature) and spintronics (by exploiting the spin of one or more interacting electrons). These simulations will be performed during the talk to show that affordable parallel scientific computing is really at our reach.