Relative Content

Tag Archive for performanceparallel-processingdijkstra

Parallel implementation of Dijkstra : problem scaling with constant efficiency

I am learning about parallelization, but I am stuck on this problem : assume that you achieve a parallel efficiency of 70% for computing the shortest path for a graph with n = 20000 vertices on 16 processors and assume that you want to use a system with 32 processors. For how many vertices you will obtain the same parallel efficiency?