$$ What strategy did you use to partition the datasets between the worker threads?

$$ How many local indexes does each worker thread create and how many times

do they update the global index with the local indexes?

$$ What mutual exclusion access solution did you use to make sure that the worker

threads update the global index in a safe fashion?

$$ How fast is your program running over Dataset$5$ when configured with $8$

worker threads versus when configured with $1$ worker thread? Explain why

your program runs faster, or slower or the same.