These operations can occur in parallel. Especially effective if you can filter on a column that is used to partition your dataset.