Adaptive deadlock-free routing mechanisms are required to handle variable traffic patterns in dragonfly networks. However, distance-based deadlock avoidance mechanisms typically employed in Dragonflies increase the router cost and complexity as a function of the maximum allowed path length. This paper presents on-the-fly adaptive routing (OFAR), a routing/flow-control scheme that decouples the routing and the deadlock avoidance mechanisms. OFAR allows for in-transit adaptive routing with local and global misrouting, without imposing dependencies between virtual channels, and relying on a deadlock-free escape subnetwork to avoid deadlock. This model lowers latency, increases throughput, and adapts faster to transient traffic than previously ...
Dragonfly networks are composed of interconnected groups of routers. Adaptive routing allows packets...
Abstract- Second generation multicomputers use wormhole routing, allowing a very low channel setup t...
In previous papers, a theory for the design of deadlock-free adaptive routing algorithms as well as ...
Abstract—Dragonfly networks are appealing topologies for large-scale Datacenter and HPC networks, th...
Dragonfly networks are appealing topologies for large-scale Data center and HPC networks, that provi...
High-radix hierarchical networks are cost-effective topologies for large scale computers. In such ne...
Abstract—High-radix hierarchical networks are cost-effective topologies for large scale computers. I...
The Cray Cascade architecture uses Dragonfly as its interconnect topology and employs a globally ada...
Evolving technology and increasing pin-bandwidth moti-vate the use of high-radix routers to reduce t...
Dragonfly networks arrange network routers in a two-level hierarchy, providing a competitive cost-pe...
Accurately estimating congestion for proper global adaptive routing decisions (i.e., determine wheth...
Abstract—Dragonfly topologies are one of the most promising interconnect designs for enabling large,...
Recently, the use of graph-based network topologies has been proposed as an alternative to tradition...
Dragonfly topologies are recent network designs that are considered one of the most promising interc...
Current HPC and datacenter networks rely on large-radix routers. Hamming graphs (Cartesian products ...
Dragonfly networks are composed of interconnected groups of routers. Adaptive routing allows packets...
Abstract- Second generation multicomputers use wormhole routing, allowing a very low channel setup t...
In previous papers, a theory for the design of deadlock-free adaptive routing algorithms as well as ...
Abstract—Dragonfly networks are appealing topologies for large-scale Datacenter and HPC networks, th...
Dragonfly networks are appealing topologies for large-scale Data center and HPC networks, that provi...
High-radix hierarchical networks are cost-effective topologies for large scale computers. In such ne...
Abstract—High-radix hierarchical networks are cost-effective topologies for large scale computers. I...
The Cray Cascade architecture uses Dragonfly as its interconnect topology and employs a globally ada...
Evolving technology and increasing pin-bandwidth moti-vate the use of high-radix routers to reduce t...
Dragonfly networks arrange network routers in a two-level hierarchy, providing a competitive cost-pe...
Accurately estimating congestion for proper global adaptive routing decisions (i.e., determine wheth...
Abstract—Dragonfly topologies are one of the most promising interconnect designs for enabling large,...
Recently, the use of graph-based network topologies has been proposed as an alternative to tradition...
Dragonfly topologies are recent network designs that are considered one of the most promising interc...
Current HPC and datacenter networks rely on large-radix routers. Hamming graphs (Cartesian products ...
Dragonfly networks are composed of interconnected groups of routers. Adaptive routing allows packets...
Abstract- Second generation multicomputers use wormhole routing, allowing a very low channel setup t...
In previous papers, a theory for the design of deadlock-free adaptive routing algorithms as well as ...