In order to design an interconnection network, it is essential to have a comprehensive understanding about properties and limitations of the network. These properties and limitations are characterized by the topology of the network. Since a topology sets constraints and costs, it plays a critical role in all interconnection networks. Different topologies have been proposed for interconnection networks in literature. The Generalized Hypercube is one of the oldest topologies that can be mentioned. Recently a group of researchers at HP Lab have introduced a new topology for these networks, called HyperX. Despite of many similarities between these two topologies, there are significant differences between their performances and costs. It seems that this important issue has been neglected in contexts of interconnection networks. In this paper, we compare HyperX and Generalized Hypercube topologies under some key topological measures. We show that HyperX is somehow better than Generalized Hypercube in the sense of topological properties.
