Patterns for High Performance Multiscale Computing

Abstract

© 2018 The Authors. We describe our Multiscale Computing Patterns software for High Performance Multiscale Computing.Following a short review of Multiscale Computing Patterns, this paper introduces the Multiscale Com-puting Patterns Software, which consists of description, optimisation and execution components. First,the description component translates the task graph, representing a multiscale simulation, to a particulartype of multiscale computing pattern. Second, the optimisation component selects and applies algorithmsto find the most suitable mapping between submodels and available HPC resources. Third, the executioncomponent which a middleware layer maps submodels to the number and type of physical resourcesbased on the suggestions emanating from the optimisation part together with infrastructure-specificmetrics such as queueing time and resource availability. The main purpose of the Multiscale ComputingPatterns software isto leverage the Multiscale Computing Patterns to simplify and automate the executionof complex multiscale simulations on high performance computers, and to provide both application-specific and pattern-specific performance optimisation. We test the performance and the resource usagefor three multiscale models, which are expressed in terms of two Multiscale Computing Patterns. In doingso, we demonstrate how the software automates resource selection and load balancing, and deliversperformance benefits from both the end-user and the HPC system level perspectives.