While in SAP HANA, both Row-based data storage and Column-based data storage is used, Column-based data storage is widely used for transactional data upon which analytical applications are implemented. This is because in Column-based data storage information is juxtaposed next to each other making access quicker. This makes aggregate functions perform better in real-time.

Apart from Quicker queries, Column-based data storage allows a higher degree of compression by virtue of common data. Majority of the columns will have a smaller set of distinct values. Consider a system with one million POs. They will have only a few hundred vendors, a dozen PO types and limited values of valid statuses. While PO numbers are unique, a number of fields will be repeated multiple times. These repetitive data is handled by attributes tables making compression possible to a higher degree.

Parallel Processing on the same set of data is rapid when operations are happening on different columns. If different operations are needed for different columns, they can be handled by different cores at the same time. At a higher level, this enables greater turn-around and increases efficiency.

~S