The analytical solution is using Fourier series to discretize the result. Technically, it has been implemented with the MPFR lib to be able to manipulate high precision numbers. It is necessary because of the calls to sinh which generated values which are not storeable in a long double

Usage:

# N: Edge size of the matrix
./laplace_analytical <N>

The results are based on 5-points finite different numerical solution to Laplace equation.

Usage:

# N: Edge size of the matrix
# w: Init value
# e: Tolerance to check convergence
./laplace_serial <N> <w> <e>

Build:

cd mpi
make -B

Usage:

# N: Edge size of the matrix
# w: Init value
# e: Tolerance to check convergence
./mpi_gs_redblack_slice_bin <N> <w> <e>
./mpi_jacobi_redblack_slice_bin <N> <w> <e>
./mpi_gs_slice_bin <N> <w> <e>

Jacobi and domain horizontal slicing Red-Black Ordering Jacobi and domain horizontal slicing