The author reports on an implementation of a finite-element code for nonlinear three-dimensional low-frequency magnetic field calculation on Intel's iPSC/2. The approach used exploits the natural parallelism in the finite-element method via parallel processing. Element assembly and the linear equation solution are the two subtasks that consume the majority of CPU cycles and are the two areas where parallelism is best exploited
Published in:
Magnetics, IEEE Transactions on
(Volume:26
,
Issue:
2
)
Date of Publication: Mar 1990