Take a CPU that shows two parallel fetch-implement pipelines for superscalar processing. Determine the performance improvement over scalar pipeline processing and no-pipeline processing, suppose an instruction cycle similar in the commentary, i.e.:
- a one clock cycle fetch
- a one clock cycle decode
- a two clock cycle execute
and an 80 instruction sequence.