Multiple jobs can run in parallel and finish faster than if they had run sequentially. Consider three jobs, each of which needs $10$ minutes of CPU time. For sequential execution, the next

one starts immediately on completion of the previous one. For parallel execution, they start to run simultaneously. In addition, "running in parallel" means that you can use the utilization

formula $($util $=1-pn)$ that was discussed in the chapter $2$ notes related to Figure $2-6.$

Guidance: for figuring completion time, consider the statements about $"$X$\%$ CPU utilization". If a job requires $($say$)10$ minutes of CPU time, then for utilization below $100\%,$ it should stand

to reason that it'll run for more than $10$ minutes of clock time. Example: $20\%$ utilization means the process only uses $12$ seconds of CPU time per wall clock minute $(12/60=20\%).$ This

also means that as a sole occupant of a CPU, it will run for $\frac{10}{0\mathrm{.}2}=50$ minutes to complete.

Given this, you can figure out total completion time for sequential and parallel jobs. For processes executing sequentially, determine how long one will spend in the absence of

competition $-$ then it's just sequential $($add their times up$).$ For parallel processes, once you figure out the total CPU utilization for $n$ processes, then each job gets an equal fraction of that.

For this problem, each job requires $15$ minutes of CPU time in total.

What is the completion time of the last one if they run sequentially, each with $50\%$ CPU utilization $(50\%$ I$/$O wait$)?$

What is the completion time of the last one if they run sequentially, each with $20\%$ CPU utilization $(80\%$ I$/$O wait$)?$

What is the completion time if they run in parallel, each with $60\%$ CPU utilization $(40\%$ I$/$O wait$)?$

What is the completion time if they run in parallel, each with $30\%$ CPU utilization $(70\%$ I$/$O wait$)?$