Faster GPU-based convolutional gridding via thread coarsening