Efficiency of ARM processors for classical molecular dynamics

Supercomputing of the exascale era is inevitably limited by power efficiency. Nowadays, different CPU architectures are considered as possible choices for these purposes. Recently, the development of ARM processors has come to the point when their floating point performance can be seriously considered for a range of scientific applications. In this paper, we present an analysis of the floating point performance of the latest ARM cores and their efficiency for the algorithms of classical molecular dynamics.