Why is correct hot/warm data classification critical when using an SAP HANA extension node?

In an SAP HANA system with persistent memory, how is data split between DRAM and persistent memory?

In SAP HANA NSE, what is the difference between column-loadable and page-loadable data, and how do you set it?

When modifying persistent-memory settings with ALTER TABLE in SAP HANA, what do the IMMEDIATE, DEFERRED, and CASCADE options do?

Which worker group and host subrole identify an SAP HANA extension node, and why does the worker group matter?

What monitoring views provide buffer-cache statistics in SAP HANA NSE, and how can you identify page-loadable tables?

What are the SAP HANA warm-data tiering options versus the cold-data options?