On the influence of the selection function on the performance of fat-trees

Authors:
F. Gilabert;M. E. Gómez;P. López;J. Duato
Affiliations:
Dept. of Computer Engineering, Universidad Politécnica de Valencia;Dept. of Computer Engineering, Universidad Politécnica de Valencia;Dept. of Computer Engineering, Universidad Politécnica de Valencia;Dept. of Computer Engineering, Universidad Politécnica de Valencia
Venue:
Euro-Par'06 Proceedings of the 12th international conference on Parallel Processing
Year:
2006

Citing 9
Cited 1

An Optimal Shortest-Path Routing Policy for Network Computers with Regular Mesh-Connected Topologies

IEEE Transactions on Computers
The turn model for adaptive routing

Journal of the ACM (JACM)
Impact of selection functions on routing algorithm performance in multicomputer networks

ICS '97 Proceedings of the 11th international conference on Supercomputing
Interconnection Networks: An Engineering Approach

Interconnection Networks: An Engineering Approach
Deadlock-Free Adaptive Routing in Multicomputer Networks Using Virtual Channels

IEEE Transactions on Parallel and Distributed Systems
k -ary n -trees: High Performance Networks for Massively Parallel Architectures

IPPS '97 Proceedings of the 11th International Symposium on Parallel Processing
On the Influence of the Selection Function on the Performance of Networks of Workstations

ISHPC '00 Proceedings of the Third International Symposium on High Performance Computing
Improving the Efficiency of Virtual Channels with Time-Dependent Selection Functions

PARLE '92 Proceedings of the 4th International PARLE Conference on Parallel Architectures and Languages Europe
Improving Routing Performance in Myrinet Networks

IPDPS '00 Proceedings of the 14th International Symposium on Parallel and Distributed Processing

Towards an efficient fat-tree like topology

Euro-Par'12 Proceedings of the 18th international conference on Parallel Processing

Quantified Score

Hi-index	0.00

Visualization

Abstract

Fat-tree topology has become very popular among switch manufacturers. Routing in fat-trees is composed of two phases, an adaptive upwards phase, and a deterministic downwards phase. The unique downwards path to the destination depends on the switch that has been reached in the upwards phase. As adaptive routing is used in the ascending phase, several output ports are possible at each switch and the final choice depends on the selection function. The impact of the selection function on performance has been previously studied for direct networks and has not resulted to be very important. In fat-trees, the decisions made in the upwards phase by the selection function can be critical, since it determines the switch reached in the upwards phase, and therefore the unique downwards path to the destination. In this paper, we analyze the effect of the selection function on fat-trees. Several selection functions are defined, compared and evaluated. The evaluation shows that selection function has a great impact on fat-trees.