Most high performance scientific applications require efficient support for collective communication. Point-to-point message-passing communication in current generation clusters are based on Send/Recv communication model. Collective communication operations built on top of such point-to-point message-passing operations might achieve suboptimal performance. VIA and the emerging InfiniBand support remote DMA operations. Such operations not only allow data to be moved between the nodes with low overhead, they also allow to create and provide a logical shared memory address space across the nodes. This feature is valuable for implementing high performance and scalable collective operations. In this paper, we focus on barrier, probably the most frequently used collective operation. We demonstrate how RDMA write operations can be used to support inter-node barrier in a cluster with SMP nodes. We also propose a scheme to exploit the shared memory within a SMP node to support fast intra-node barrier. Combining these two schemes, we developed an overall barrier scheme for cluster of SMP nodes with cLAN VIA interconnect.Compared to the current barrier algorithms using send/recv communication model, the new approach is shown to reduce barrier latency on a 64 processor (32 dual nodes) system by up to 66 percent.