Scalability of Full Wave and Asymptotic Solvers in Altair Feko

Electromagnetic simulation time and memory are dependent upon the electrical size of the model being simulated. Electrically large models demand higher memory and runtimes to simulate and produce accurate results. Methods to remedy this requirement include using alternative computational electromagnetic (CEM) solvers to perform the analysis or increasing memory and the number of parallel processes on the CPU for the simulation. Altair Feko has various CEM solvers available at its disposal to provide options for a user to analyze electrically large and complex models at higher frequencies. Feko can also scale in parallel processes which may aid in reducing the runtime requirements of a simulation. In this paper, analysis between simulation runtime and memory between Feko’s full wave/asymptotic solvers and parallel processes is performed using aircraft models provided in UT Austin’s RCS Benchmark Suite.