Revisiting the allocation of regular data arrays to a mesh of processors | IEEE Conference Publication | IEEE Xplore