Detail publikace

Communication Performance of Mesh- and Ring-Based NoCs

Originální název

Communication Performance of Mesh- and Ring-Based NoCs

Anglický název

Communication Performance of Mesh- and Ring-Based NoCs

Jazyk

en

Originální abstrakt

As multi-core systems begin to appear, their possible applications, parallel performance and on-chip interconnection networks have to be clarified, analyzed and optimized. The paper investigates an impact of collective communication (CC) overhead   that may be critical for performance of parallel applications. Two potential topologies of networks on chip (NoC) are investigated, a ring-based network and 2D-mesh, due to their easy manufacturability on a chip. The wormhole switching, full duplex links and 1-port non-combining as well as combining nodes are considered. The lower bounds on the number of communication steps and upper bounds of CC times based on real CC algorithms are given. They can be evaluated for any given start-up time and link bandwidth. This enables performance prediction of applications with CCs among computing nodes.

Anglický abstrakt

As multi-core systems begin to appear, their possible applications, parallel performance and on-chip interconnection networks have to be clarified, analyzed and optimized. The paper investigates an impact of collective communication (CC) overhead   that may be critical for performance of parallel applications. Two potential topologies of networks on chip (NoC) are investigated, a ring-based network and 2D-mesh, due to their easy manufacturability on a chip. The wormhole switching, full duplex links and 1-port non-combining as well as combining nodes are considered. The lower bounds on the number of communication steps and upper bounds of CC times based on real CC algorithms are given. They can be evaluated for any given start-up time and link bandwidth. This enables performance prediction of applications with CCs among computing nodes.

BibTex


@inproceedings{BUT27603,
  author="Václav {Dvořák}",
  title="Communication Performance of Mesh- and Ring-Based NoCs",
  annote="As multi-core systems begin to appear, their possible applications, parallel
performance and on-chip interconnection networks have to be clarified, analyzed
and optimized. The paper investigates an impact of collective communication (CC)
overhead   that may be critical for performance of parallel applications. Two
potential topologies of networks on chip (NoC) are investigated, a ring-based
network and 2D-mesh, due to their easy manufacturability on a chip. The wormhole
switching, full duplex links and 1-port non-combining as well as combining nodes
are considered. The lower bounds on the number of communication steps and upper
bounds of CC times based on real CC algorithms are given. They can be evaluated
for any given start-up time and link bandwidth. This enables performance
prediction of applications with CCs among computing nodes.",
  address="IEEE Computer Society",
  booktitle="Proceedings of the 7th Int. Conference on Networking",
  chapter="27603",
  edition="NEUVEDEN",
  howpublished="print",
  institution="IEEE Computer Society",
  year="2008",
  month="april",
  pages="156--161",
  publisher="IEEE Computer Society",
  type="conference paper"
}