Parallel Distributed Breadth First Search on the Kepler Architecture

Mauro Bisson, Massimo Bernaschi, Enrico Mastrostefano
Istituto per le Applicazioni del Calcolo, IAC-CNR, Rome, Italy
arXiv:1408.1605 [cs.DC], (7 Aug 2014)


   author={Bisson, Mauro and Bernaschi, Massimo and Mastrostefano, Enrico},

   title={Parallel Distributed Breadth First Search on the Kepler Architecture},

   journal={ArXiv e-prints},




   keywords={Distributed, Parallel, and Cluster Computing},




Download Download (PDF)   View View   Source Source   



We present the results obtained by using an evolution of our CUDA-based solution for the exploration, via a Breadth First Search, of large graphs. This latest version exploits at its best the features of the Kepler architecture and relies on a 2D decomposition of the adjacency matrix to reduce the number of communications among the GPUs. The final result is a code that can visit 400 billion edges in a second by using a cluster equipped with 4096 Tesla K20X GPUs.
No votes yet.
Please wait...

* * *

* * *

HGPU group © 2010-2021 hgpu.org

All rights belong to the respective authors

Contact us: