Scaling of Distributed Multi-simulations on Multi-core Clusters

Abstract

DACCOSIM is a multi-simulation environment for continuous time systems, relying on FMI standard, making easy the design of a multi-simulation graph, and specially developed for multi-core PC clusters, in order to achieve speedup and size up. However, the distribution of the simulation graph remains complex and is still the responsibility of the simulation developer. This paper introduces DACCOSIM parallel and distributed architecture, and our strategies to achieve efficient multi-simulation graph distribution on multi-core clusters. Some performance experiments on two clusters, running up to 81 simulation components (FMU) and using up to 16 multi-core computing nodes, are shown. Performances measured on our faster cluster exhibit a good scalability, but some limitations of current DACCOSIM implementation are discussed.

Toward an accurate and fast hybrid multi-simulation with the FMI-CS standard

“…But despite optimization, the current FMI-CS features can lead to numerous rollbacks and reruns of FMU computations to process unexpected events. So, such a solution could limit the performance and scalability achieved on multi-core PC clusters by our pure FMI-based distributed multi-simulations [6]. Moreover, discrete signals often used by cyber-physical components in electrical grids cannot be well addressed by this strategy.…”

“…These components consume very few CPU time (fast control algorithms) but increase communications (reading many sensor outputs and sending command signals). So, this evolution should impact FMU distribution algorithms on PC clusters, which optimize both load balancing and inter-node communications [6].…”

“…Until now the FMI-CS standard allowed us to compute different kinds of Smart Grid simulation from building thermics [6] to MV network voltage management based on stateestimation. These simple use cases permitted us to identify the limits of the current standard and design new ways to overcome them.…”

On the parallelization and performance analysis of Barnes–Hut algorithm using Java parallel platforms

“…In the modern era, multi-core processors render the feasibility in attaining the higher performance in terms of execution times through efficient distribution of workload [1]. However, if the applications running on multi-core processors are not parallelized, they might produce unsatisfactory results adhering to the huge time complexity than the applications running on a single-core processor.…”

FMU-proxy: A Framework for Distributed Access to Functional Mock-up Units

“…Of these tools, four support distributed invocation of FMUs. These are: DACCOSIM (Distributed Architecture for Controlled CO-SIMulation) (Galtier et al, 2015;Dad et al, 2016), a FMI compatible master algorithm, that lets the user design and execute a simulation requiring the collaboration of multiple FMUs on multi-core computation nodes or clusters. DACCOSIM is implemented in Java and is built on-top of the Eclipse Rich Client Platform, which provides the user with a GUI for setting up and running co-simulations.…”