You are writing some parallel code to run simulations for a research project under your advisor. Suppose axcu!! $"$

the code initially starts out as a single threaded process and performs some initial setup. The program then

executes in parallel. When run sequentially, the initial setup time is $12\mathrm{.}5\%$ of the overall execution time.

$($a$)$ Your advisor wants the code to run as fast as parallel and has access to Rockfish, a supercomputing cluster

at Hopkins. What is the maximum theoretical speedup for a single simulation?

$($b$)$ Now suppose that there are $1024$ simulations that your advisor would like to run. Assume that the sim$-$

ulation code is only run sequentially. Each simulations takes the same amount of time to complete, but

multiple simulations can run simultaneously. What is the maximum speedup?

$($c$)$ Suppose you run your simulation code in parallel and run multiple simulations simultaneously. What is

the maximum speedup? Amdahls law may be usefull