The big idea: code pushdown
Classic ABAP often followed a simple pattern: read data from the database into internal tables, then process it in application code. HANA turns this around — because the database is extremely fast, it is often better to push the work down to the database and bring back only the result.
This shift is called code pushdown, and it is the single most important idea behind ABAP on HANA.
CDS views: modelling data in the database
Core Data Services (CDS) views let you define rich data models directly in the database layer, with joins, aggregations and annotations. Instead of fetching rows and looping over them, you describe the result you want and let HANA compute it.
CDS views are also the foundation of modern SAP development: they feed OData services, analytical apps and the RAP programming model.
AMDP: database procedures from ABAP
ABAP Managed Database Procedures (AMDP) let you write database-side logic in SQLScript while staying inside the ABAP world — the procedure is defined in an ABAP class, transported like any other ABAP object, and called from ABAP code.
What stays the same
The fundamentals did not change. Data Dictionary knowledge, internal tables, modularization, debugging and clean code are as valuable on HANA as they ever were. ABAP on HANA is an extension of classic ABAP, not a replacement for it.