The advent of 32-bit busses which support full 32-bit microprocessors has signaled a coming of age of microcomputer technology for simulation applications. The VME-bus, a fully non-proprietary bus standard developed through a cooperative effort of major semiconductor manufacturers, has board-level product offerings from over 200 vendors. As this number continues to grow, so does the competitive base of the various components (in low cost, general purpose form) needed to build high fidelity simulators. Although the growing availability and variety of these off-the-shelf components do not minimize the need for sound system integration, they can facilitate this crucial process, and more importantly, afford an opportunity to rethink how it can be done better. For example, in the past, shared memory has been widely used to communicate between multiple cpu's in a training system. Should this concept be adopted in microprocessor-based systems, or will emerging message passing and data broadcasting methods serve our needs more effectively? Once the best communications mechanism has been selected, what intelligence will be needed to partition the simulation, determine the data that must be transferred, and synchronize the distributed software modules of the system such that the fidelity of the simulation math model is preserved? Research has been undertaken both within the government and by military contractors to answer these questions. The Systems and Computer Technology Division of NTSC has contributed through the building of an advanced development model based on the VME-bus and commercially available components.