The next chapter covers optimizing individual programs within partitions. Once you have parallelized the pipeline (this chapter), tuning each partition's program (Chapter 26) provides additional elapsed time reduction.