I am a DB2 DBA for z/OS, but I have been learning the LUW world lately and I have a question about partitioning. On z/OS partitioning is at the table level, and I can have partitions that are 64GB each, where a non-partitioned tablespace can only be 64GB. I know there is something called database partitioning and range partitioning (table) in LUW but I am trying to find limits on the table partitioning. I know (at least I think I know) the max size on a large tablespace on LUW 9.5 is 16TB, but if I partition it by range does that mean I get 16TB x # of partitions? Or does the database partitioning better relate to the z/OS partitioning? And if anyone what's to play teacher, what would be the reasons to use database partitioning or table/range partitioning?
Thanks!

Yes. For your other questions: read this DB2 Diary from Radhesh: DB2 Partitioning

Maximum tablespace size depends on its page size, so the number will be different for each page size tablespace. There is also a limit on the table partition size. Please refer to the SQL and XML Limits section in the manual. SQL and XML limits

Briefly, database partitioning (DPF) allows you to parallelize queries and utility execution, while table partitioning is good for data maintenance (archiving etc.) and range-based queries (if they map to the partitioning key, that is).

Table Partitioning in DB2 LUW is equivilent to Range Partitioning in DB2 z/OS.

Database partitioning (DPF) is a whole 'nother animal and is used for MPP (Massive Parallel Processing). Data is randomly distributed accross multiple logical (or physical) nodes based on a hash key. The closest thing to DPF is Teradata. DPF requires extra cost licensing over and above DB2 ESE edition. Starting in 9.7 they renamed DPF feature as InfoSphere Warehouse Edition (there are actually several different InfoSphere Warehouse editions such as Basic, Enterprise, etc).