When discussing parallel processing performance, the terms speedup and scalability often arise. This chapter describes how you can measure and improve speedup or scalability when running with either shared memory or distributed memory parallel activated, including the use of GPUs. The following topics are available: