Previous | Next --- Slide 6 of 52
Back to Lecture Thumbnails

Speedup is the time it takes to run on 1 processor divided by the time it takes to run on n processors.

The fraction f is the parallel portion of the program and it would take time f/n because the work can be parallelized over n cores. The two assumptions here is that work is distributed evenly across the processors (perfect parallelism) and that all the processors are the same (all have same performance). The rest of the time, (1 - f), is for doing the sequential part of the work.

Note that if the value of f is too small, there won't really be much/any parallel speedup.