Code scaling |
| |
Authors: | Schuh M.J. Woo A.C. |
| |
Affiliation: | NASA Ames Res. Center, Moffett Field, CA; |
| |
Abstract: | The order of an electromagnetics algorithm or code is defined by the rate at which the CPU time, and memory requirements CCPU=O(fα) and Cm=O(fβ ), respectively-grow with frequency. Knowledge of this information helps in the prediction of computer-run times and memory requirements for problems of interest. This paper presents a methodology for determining the computational and memory orders of a code. These results are presented for a finite-volume time-domain (FVTD) code and for a method-of-moments (MoM) code. Computer-resource requirements are plotted for computer runs that calculate the bistatic radar cross section (RCS) for spheres of different electrical sizes, within a set error level from the Mie-series solution. These plots are used to calculate the computational and memory orders of the codes |
| |
Keywords: | |
|
|