An in-depth study of dimension-extended dragonfly interconnection network

Dragonfly topology is a commonly utilized design for interconnection networks in parallel and distributed systems. A classical dragonfly can be denoted as dragonfly($k$,$m$,$l$), where$m$ is the number of routers in a group,$l$ is the number of links per router connected to other groups, and$k$ is the number of links per router connected to compute nodes. Each router has other$m-1$ links fully connected to other$m-1$ routers within a group. Each group has$ml$ links connected to other groups. The groups are also fully connected, therefore there are$ml+1$ groups in total. The router radix in a dragonfly($k$,$m$,$l$) is$l+k+m-1$. Building a large dragonfly system requires a large number of high-radix routers, increasing hardware costs. To reduce hardware costs, this paper proposes a more flexible topology called dimension-extended dragonfly (DED). Rather than routers in a group being fully connected, each router in a group is arranged in an$n$-dimensional matrix, and routers of the same dimension are fully connected. We use$n$ to denote the dimension such that each group in the DED has$m^n$ routers. This study comprehensively evaluates DED in terms of cost, performance, fault tolerance, and packet latency. The findings show that DED provides a more economical hardware solution compared to traditional Dragonfly and Cascade topologies, especially for$n\ge 3$. Beyond cost-efficiency, DED enhances system design flexibility. It offers diverse possibilities for system scaling through different combinations of diameter and radix, giving system architects more adaptable options. To further enhance the versatility of DED, three disjoint path routing algorithms are proposed and their fault tolerance is evaluated through simulation. The simulation results also show that the packet latency of DED is lower than dragonfly and cascade.