The Operational Environment (OE) has become increasingly complex, with challenging human factors, exponential proliferation of technology, and an increasingly determined, adaptive threat. Training Army Soldiers, leaders and units in a complex world requires modernized, integrated, realistic, and adaptive training capabilities. The Army must leverage emerging technologies to transform the way it develops and delivers training to enable agile and adaptive Soldiers, leaders and versatile units. It must provide a consistent, persistent ability to train at the point of need (PON) for both current and future operations as part of a Joint, Inter-organizational, and Multinational (JIM) force. The training venues must allow the Army to train as it fights, using its wartime systems on its operational networks and all training environments must replicate the OE to the greatest extent possible. To address this reality, the U.S. Army Program Executive Office for Simulation, Training and Instrumentation (PEO STRI) and the University of Central Florida (UCF), Institute for Simulation and Training (IST) are conducting research to create new capabilities that support both operations and training while enabling software application migration to Army enterprise data centers and cloud environments. This research pivots on Army directives that focus on modernizing information technology systems and applications. To achieve the distributed nature of this vision, technical enhancements to the underlying Army Enterprise Network (AEN) must be made in the next few years. This paper investigates potential gaps in simulation enterprise network architectures and describes research results in three major technical areas that address these gaps and will benefit future simulation and network architectures. The research topics include technologies that: (1) efficiently delivers simulations from cloud-like environments using Software Defined Networks (SDNs); (2) facilitates individual and collective home station or field-based training through the use of thin clients; and (3) optimizes computational resources through load-balancing techniques and processes.