Comparison of Cloud and Local HPC Approach for MUSCLE-based Multiscale Simulations

Abstract

In this paper we present and compare a support for setting up and execution of multiscale applications in the two types of infrastructures: local HPC cluster and Amazon AWS cloud solutions. We focus on applications based on the MUSCLE framework, where distributed single scale modules running concurrently form one multiscale application. We also integrate presented solution with Grid Space virtual laboratory that enables users to develop and execute virtual experiments on the underlying computational and storage resources through its website based interface. Last but not least, we present a design of a user friendly visual tool supporting application distribution.