D. Lázaro Cuadrado, A.P. Ravn, and P. Koch (Denmark)
Distributed processing, automatic deployment, simulation, actor, Ptolemy, speedup.
Despite the well known advantages of distributed processing for intensive computations like simulation, frameworks often fail to exploit them. A distributed simulation is harder to develop than a sequential one, because it is necessary to interface and map activities to processors and handle the ensuing communication and synchronization problems. Very often the designer has to explicitly specify extra information concerning distribution for the framework to make an effort to exploit parallelism. This paper presents Automated Distributed Simulation (ADS), which allows the designer to forget about distribution concerns while benefiting from the advantages. ADS relies on the actor formalism. It is realized as an open source implementation for the Ptolemy II simulation framework. Experiments compare different topologies, granularities and number of blocks, achieving linear speedups for practical cases. We implement pipelining techniques so iterative models with purely sequential topologies can benefit from ADS.
Important Links:
Go Back
