Start by recalling the sequential work and span complexities for matrix addition and multiplication. How do these translate to parallel settings?

For matrix addition, analyze how dividing the work across processors affects the span. What is the critical path length in terms of the matrix dimensions?

For matrix multiplication, consider the parallelization strategies (e.g., Cannon’s algorithm or Strassen’s). How does the span depend on the choice of algorithm, and what is the work distribution?