- Amdahl's Law describes the speedup from parallelizing parts of a computation. It shows that the overall speedup is limited by the time spent on sequential parts of the program. - Two examples show calculating the maximum speedup from parallelizing 90% and 25% of a program's operations using multiple processors. - Pipelining a CPU with a 5-stage pipeline and lower clock speed from 2.5GHz to 2GHz achieves a speedup of 3.2x due to reducing the time between instructions. Pipelining a 11-stage processor sees a speedup of 9.167x for executing 50 instructions.