Computational advances in gravitational microlensing: a comparison of CPU, GPU, and parallel, large data codes