Finite-difference time-___domain method: Difference between revisions

Revision as of 02:32, 6 July 2025

Revision as of 03:22, 27 July 2025

Line 175:

# Parallel-processing computer architectures have come to dominate supercomputing. FDTD scales with high efficiency on parallel-processing CPU-based computers, and extremely well on recently developed GPU-based accelerator technology.<ref name="taflove05" />

# Computer visualization capabilities are increasing rapidly. While this trend positively influences all numerical techniques, it is of particular advantage to FDTD methods, which generate time-marched arrays of field quantities suitable for use in color videos to illustrate the field dynamics.<ref name="taflove05" />

# Anisotropy is treated naturally by the FDTD method. Yee cells, having components in each Cartesian direction, can be easily configured with anisotropic characteristics.<ref name="taflove05"/>

Taflove has argued that these factors combine to suggest that FDTD will remain one of the dominant computational electrodynamics techniques (as well as potentially other [[multi-physics|multiphysics]] problems).<ref name="taflove05" />

the dominant computational electrodynamics techniques (as well as potentially other [[multi-physics|multiphysics]] problems).<ref name="taflove05" />

==See also==