You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Hi, I successfully compiled the program, process the edge data files, but failed to output the correct result.
I use the data from https://snap.stanford.edu/data/soc-pokec.html , consisting of 30 million edges, run your process text_to_bin.bin by setting reverse=1 . Then I run the gpu_ibfs.bin with conrrent_count=64(128) and bfs_count=1, because I want to see the performance for single instance. But I found that the total time elapsed is about 9*10e-7, i.e. less than 10e-6 seconds. That cannot be true, for such a graph, right?
Any suggestions ?
The text was updated successfully, but these errors were encountered:
Hi, I come to realize that src_count>= concurrent_count should be maintained. If I want to test the performance of single instance bfs, what should I do? Is it better to run Enterprise directly ?
Yes. src_count >= concurrent_count should be maintained. If you want to test single BFS performance, again, you are right, you better run Enterprise directly.
Hi, I successfully compiled the program, process the edge data files, but failed to output the correct result.
I use the data from https://snap.stanford.edu/data/soc-pokec.html , consisting of 30 million edges, run your process text_to_bin.bin by setting reverse=1 . Then I run the gpu_ibfs.bin with conrrent_count=64(128) and bfs_count=1, because I want to see the performance for single instance. But I found that the total time elapsed is about 9*10e-7, i.e. less than 10e-6 seconds. That cannot be true, for such a graph, right?
Any suggestions ?
The text was updated successfully, but these errors were encountered: