Civilian and Department of Defense (DoD) organizations are leveraging immersive simulation to create a digital twin of the world. This effort to maximize training and capability invokes difficulties when used in the field due to the processing power necessary to reconstruct photogrammetric environments in short suspense. Edge processing typically overlooks two critical areas (1) the hardware best suited to accomplish the task and (2) the most advantageous strategy for integration between the processing system and the software which will power the experience. With optimization in these areas, significant decreases in processing time can be achieved while decreasing the size, weight, and power consumption (SWaP) of the system. This allows for man-portable systems which can be employed directly with the user for training and operational use. 

This paper will address the issues stated above by sourcing firsthand data through hardware and photogrammetric software processing benchmarks across three system configurations; multi-socket single node, single socket dual node, and single socket quad node. This data will then be aggregated into the developmental foundation for a new system topology which will offer server grade reconstruction speeds in a workstation form factor. This system will showcase the effects of hardware optimization through decreased processing times, weight, power consumption, and heat generation. A real-world scenario will illustrate the mission-critical system performance and its potential infusion into the One World Terrain ecosystem. Finally, the real-world benefit of focusing on these key components in simulation-based systems for training will be explained along with courses of action for implementation by warfighters at the edge. Converging the advancements in software architecture with the evolution of central processing units (CPU) and graphics processing units (GPU), predicted by Moore’s Law, we will set the stage for the next epoch of photogrammetric systems and software. 

Keywords
CHARACTERIZING SYSTEM PERFORMANCE ,CONTENT GENERATION,DENIED ENVIRONMENTS,INTELLIGENCE,TRAINING

Additional Keywords
Photogrammetry, Digital Twin, Edge Computing, Simulation & Modeling,