Various topologies have been proposed for High Performance Computing(HPC). i.e., fat-tree, Torus topology. Compared with conventional fat-tree topology, torus performs much better when applied in HPC. Unfortunately, due to its wraparound links, Torus topology naturally has a tendency to trigger deadlock incidents inside the network. People solve this problem by means of virtual channel, but this approach will also restrict the routing of message. In this paper, we propose a deadlock-free topology for HPC, without the utilization of virtual channel approach, called Mesh-of-Torus, which incarnates the good characteristics of mesh and torus topology. The new topology possesses the same network diameter as the torus topology does. Furthermore, we have proposed a corresponding port assignment mechanism in consideration of deadlock avoidance without the employment of virtual channel, complicated internal arbitration and scheduling mechanism when dimension-order routing algorithm is employed. The switching fabric of server in Mesh-of-Torus is simplified when compared with classical torus topology. Finally, simulations and mathematical analysis have shown that Mesh-of-Torus outperforms hierarchical mesh and traditional torus in terms of average end-to-end latency and network load distribution.