
2.5 Exercises 117
2.5 EXERCISES
2.1 Describe two examples where a good parallel algorithm must be based on a serial
algorithm that is different from the best serial algorithm since the latter does not
afford enough concurrency
2.2 Which of our case study applications (Ocean, Barnes-Hut, Raytrace, and Data
Mining) do you think are amenable to decomposing data rather than computation
and using an owner computes rule in parallelization? What do you think the prob-
lem(s) would be with using a strict data distribution and owner computes rule in
the others?
2.3 There are two dominant models for how parent and children processes relate to
each othe ...