Shortest path algorithms for different applications, such as Internet routing, VLSI design and so on are used. Dijkstra and Bellman-Ford are commonly used shortest path algorithms which are typically implemented in networks with hundreds of nodes. However, scale of shortest path problems is increasing, and more efficient algorithms are needed. With the development of multicore processors, one natural way to speedup shortest path algorithms is through parallelization. In this paper, we propose a novel shortest path algorithm with parallel vertex transversal, and compare its speed with standard solutions in datacenter topologies.