FEDS can be run on a machine connected to a network. In some circumstances FEDS can even be run remotely from a network computer or file share. However, because each network is different and interference and connectivity issues are possible, for the best performance (and to lower the risk of problems) it is recommended to run FEDS from a local machine.

A useful rule of thumb is that a full FEDS optimization run will take a couple minutes per building set. However, run-time depends on a number of factors, including computing resources, processor speed, size, and complexity of the case. The more buildings, use areas, and technologies being analyzed, the longer the FEDS run will take. Additionally, the presence and number of central energy plants and thermal loops will also impact run-time. 

Conversely, a calibration run of the same case and computer takes only a few seconds (with additional time to generate the reports). This is because all building sets are excluded from optimization to help focus on the baseline energy results and aid in focusing on model quality assurance and calibration processes.

FEDS contains a built-in database of building survey data and is able to infer a number of building parameters based on the small set of required inputs provided by the user. For example, FEDS uses information such as building type, location, floor area, and vintage to determine the most likely construction type and geometry. It uses similar information along with heating fuel type and cooling equipment, to determine the most likely heating technology and ventilation system parameters for a building. All inferences enable a user to model buildings without having intimate knowledge of the detailed engineering parameters. The resulting building prototype parameter values are statistically the most likely values based on the limited set of information provided. Of course, all inferred data may be easily overwritten by simply entering (locking) a value in the user interface screens.