Multigrid method with adaptive IDR-based Jacobi Smoother

The multigrid (MG) method is known as the fast linear solver for large scale finite element analyses. In the iterations of MG, the Gauss-Seidel method is usually adopted as the smoother. Recently, Induced Dimension Reduction (IDR)-based methods receive much attention because these methods solve faster than the other strong solvers such as BiCGStab method. In this paper, we present a performance of MG with adaptive IDR-based Smoother method. Numerical results show that the present method has good convergence and good parallel efficiency in the electromagnetic analyses.