T. Jacques et al., Implementation of the boundary integral method on MIMD systems for electromagnetic scattering problems, IEEE MAGNET, 36(4), 2000, pp. 1479-1483
The boundary integral method is used to solve problems of scattering by per
fect electric conducting or perfect dielectric bodies. Because this method
requires large memory storage, it is implemented on a distributed memory pa
rallel computer. The assembling is performed by nodal contribution, and the
BiCGStab(m) algorithm is used for solving. Performances are analyzed using
several large problems.