Multi-formalism modeling approach for semiconductor supply/demand networks

Building computational models of real world systems usually requires the interaction of decision modules and simulation modules. Given different models and algorithms, the major hurdles in building a principled and robust system are model composibility and algorithm interoperability. We describe an approach to the composibility problem including initial results. This exposition is given in the context of linear programming as the decision technique and discrete event simulation as the simulation technique, both applied to the design and operation of semiconductor supply/demand networks.