Topology-oblivious optimization of MPI broadcast algorithms on extreme-scale platforms
Authors:
- Khalid HasanovKHKhalid Hasanov
- Jean-Noël QuintinJQJean-Noël Quintin
- Alexey LastovetskyALAlexey Lastovetsky
- 1 November 2015
- journal article
- research article
- Published by Elsevier in Simulation Modelling Practice and Theory
- Vol. 58, 30-39
Abstract
No abstract availableKeywords
Funding Information
- Irish Research Council for Science, Engineering and Technology
- IBM (EPSPG/2011/188)
- Science Foundation Ireland (08/IN.1/I2054)
This publication has 12 references indexed in Scilit:
- High-Level Topology-Oblivious Optimization of MPI Broadcast Algorithms on Extreme-Scale PlatformsLecture Notes in Computer Science, 2014
- Designing Optimized MPI Broadcast and Allreduce for Many Integrated Core (MIC) InfiniBand Clusters2013
- Cheetah: A Framework for Scalable Hierarchical Collective Operations2011
- Optimal broadcast for fully connected processor-node networksJournal of Parallel and Distributed Computing, 2008
- The deep computing messaging framework2008
- A practically constant-time MPI Broadcast Algorithm for large-scale InfiniBand Clusters with Multicast2007
- Open MPI: Goals, Concept, and Design of a Next Generation MPI Implementation2004
- A bandwidth latency tradeoff for broadcast and reductionInformation Processing Letters, 2003
- A PIPELINED BROADCAST FOR MULTIDIMENSIONAL MESHESParallel Processing Letters, 1995
- Optimum broadcasting and personalized communication in hypercubesIEEE Transactions on Computers, 1989