Divides core input data into manageable blocks, each representing a subsection of data that is small enough to be processed by a single worker.