Hi Vivek,
The BW infocube is an extended star schema, where our fact tables contain DIMID's which connect then to dimension tables and the dimension tables in-turn contain the SID to connect to your master data tables.
A infocube on a BW on HANA system would have the SID's in the fact table itself, there by eliminating the need for a dimension table.
Why the infocube was built? HANA stores data in-memory, but the traditional RDBMS that we use does not, so it was prudent having a extended star schema but now since all data is in-memory, we dont need the over head of having a dimension table anymore.
Hope this helps.
Regards,
Benedict