Scaling Up Recommender

Let’s say we want to train a recommender system for Netflix. We can train a model with a training size of thousands of movies and users.

How do would scale the training of this model on the entire Netflix data which has millions of movies and users?