2. Suppose a sequential program S has an execution time 100 msec. Further, assume that SP is a parallel variant of S. After an experimental evaluation over some different numbers of processors, the following running times were achieved: a. For each experimental configuration, compute speed-up and, in turn, Karp-Flatt Metric. Round up your calculations to 2 decimal places. ( 2m2m) b. Explain what is your opinion regarding parallelization overhead? ( c. Suppose prior to the experimental evaluation, we computed speedup using Amdahl's Law. Would those speedup values be higher than, equal to, or lower than the speedup values computed in part (a) above? Why? ()