So 1e6 light year radius to match the solar irradiance from purely inverse square losses? Wow.