The MultiLevel fast Physical Optics (MLPO) algorithm allows to perform Physical Optics integrals with a computational complexity comparable to that of the Fast Fourier Transform (FFT). The multilevel scheme is based on hierarchical domain decomposition and phase compensated interpolation approach. An optimized distributed memory parallelized MLPO algorithm has been designed by partitioning not only the radiating aperture but also the grid of far field directions. The scalability of this algorithm on a large grid is discussed, based on a performance analysis of an MPI-parallelized code.