Difference between revisions of "Graph traversal"

From Rodinia
Jump to: navigation, search
Line 4: Line 4:
 
This benchmark suite provides the GPU implementations of breadth-first search (BFS) algorithm which traverses all the connected components in a graph as well as single source shortest path algorithm (SSSP) which finds the smallest combined weight of edges between nodes.
 
This benchmark suite provides the GPU implementations of breadth-first search (BFS) algorithm which traverses all the connected components in a graph as well as single source shortest path algorithm (SSSP) which finds the smallest combined weight of edges between nodes.
 
   
 
   
 +
We would like to acknowledge Pawan Harish and P.J.Narayanan at [http://cvit.iiit.ac.in/index.php?page=projects IIIT] who contributed their code to the Rodinia benchmark suite. The related paper can be found at [http://www.iiit.net/techreports/2007_103.pdf here]
  
 
+
*[[Downloads]]
The breadth-first search and shortest path CUDA implementations can be downloaded from [http://www.people.virginia.edu/~sc5nf/HiPC.zip here]
+
 
+
We would like to acknowledge Pawan Harish and P.J.Narayanan at [http://cvit.iiit.ac.in/index.php?page=projects IIIT] who contributed their code to the Rodinia benchmark suite. The related paper can be found at [http://www.iiit.net/techreports/2007_103.pdf here]
+

Revision as of 02:58, 28 January 2009

Graph algorithms are fundamental and widely used in many disciplines and application areas. Large graphs involving millions of vertices are common in scientific and engineering applications. This benchmark suite provides the GPU implementations of breadth-first search (BFS) algorithm which traverses all the connected components in a graph as well as single source shortest path algorithm (SSSP) which finds the smallest combined weight of edges between nodes.

We would like to acknowledge Pawan Harish and P.J.Narayanan at IIIT who contributed their code to the Rodinia benchmark suite. The related paper can be found at here