Structure of HybridProviders Based on Direct Access 
The HybridProvider on direct access is based on a combination of VirtualProvider and InfoCube. Both objects have the same structure. When you edit the HybridProvider, you edit both InfoProviders in the same way.
The data is loaded using a data transfer process that you create from the InfoCube of the DataSource. The system also creates a data transfer process between VirtualProvider and DataSource, that uses the same transformation.

The graphic below shows the data flow in a HybridProvider based on direct access:
When the query is executed on the HybridProvider, most of the data is read from the InfoCube. The most up to date data is also read from the source system using the VirtualProvider. Only the data that has not been loaded into the InfoCube is read. This read mode is named Hybrid Delta.
However the read mode Hybrid Delta is not supported by all DataSources that can be connected to a HybridProvider. In other words, you cannot always read the most up to date data from the source system due to technical reasons. However the use of a HybridProvider based on a direct access is still recommended. In this case, you use the read mode Switch. This means that the data is read either from the InfoCube or from the VirtualProvider. Using a variable, you can control what data is read while the query is running.