It's not actually in the database. It's a lousy implementation which does not use the distributed nature of HANA. You fetch the data first and then you do your analyses in R which is hosted near the DB. If you want parallel processing in R, you're on your own.
