Interacting With ArtiSynth

A visualization of the model independence

Because of the architecture of the Batch Simulation Framework ( BSF ), using it to conduct batch simulations is easy.  Virtually any model developed in ArtiSynth can be adapted to run through the framework because of the abstraction between ArtiSynth properties and the model components.  This is known as 'model-independence.'

The BSF Architecture

A visualization of the Manager-Worker architecture

The BSF adheres to the Manager-Worker hierarchy, which allows the ArtiSynth simulation to take advantage of large computing power.  By having a single manager distribute simulation tasks to any number of workers, the platform on which the program is running can perform them all in parallel.  This can easily cut down on simulation time by 10x to 20x or more.


A probability distribution

BSF simulation can run in a couple of modes.  If the simulations are organized through combinatrics, the BSF can initialize and run all combinations of input streams.  This allows the user to cover a huge range of simulation states with a relatively small input set.  By providing only 'm' sets of size 'n' input data, the BSF can run nunique simulations with ease.

Probabalistic Simulations

BSF also includes the capability of running probablistic simulations, which involve specifying a distribution set for input. It utilizes the common 'Monte-Carlo' method of probablistic sampling.