SSIS processes data in the pipeline within memory buffers to efficiently load and manipulate datasets in memory. The main advantage of managing data in memory is that it is faster than storing data physically on the disk.
In most packages, changing the SSIS buffer properties is not required, because SSIS does a very good job optimizing it. But even so, it would be possible to tune the Data Flow performance manually to change the values of the SSIS buffer properties. The size of each buffer is determined at runtime and buffers will be created as required in order to process all the data.
The number of buffer created is dependent on how many rows fit into a buffer and the number of rows fitting into a buffer depends on defined data types and the length of each column. There are three internal parameters used to calculated the number of buffers: