It already contains all the internal logic for: Data partitioning Task scheduling Parallel execution Memory management Fault tolerance When we use PySpark or Spark SQL, we are simply using an SDK to ...