#PageRank Matrix Multiplication Parallelizing raising sparse matrices to high powers with OpenMPI
CPSC 424 Parallel Programming Final Project by Amelia Holcomb and Jay Hou
- dot product
- log2 of power instead of power # of iterations
- To change columns to rows between processors, we wanted to use scatterv. But then we realized that this isn't so nice because if we assign more than one column per process, we don't know where to put the entries in the receiving process. Instead, we'll receive them out of order. To ameliorate this, we are taking it out of order, but doing a fast n*log(procs) merge-sort on the vectors.