hybrid_BC
所属分类:GPU/显卡
开发工具:Cuda
文件大小:6801KB
下载次数:0
上传日期:2018-12-20 18:31:34
上 传 者:
sh-1993
说明: GPU上并行中介中心性的混合方法
(Hybrid methods for Parallel Betweenness Centrality on the GPU)
文件列表:
LICENSE (1579, 2018-12-21)
graphs (0, 2018-12-21)
graphs\breast_coexpress.txt (2875501, 2018-12-21)
graphs\delaunay_n10.graph (24991, 2018-12-21)
graphs\delaunay_n11.graph (56686, 2018-12-21)
graphs\delaunay_n12.graph (120120, 2018-12-21)
graphs\delaunay_n13.graph (247126, 2018-12-21)
graphs\delaunay_n14.graph (539299, 2018-12-21)
graphs\delaunay_n15.graph (1145388, 2018-12-21)
graphs\delaunay_n16.graph (2358015, 2018-12-21)
graphs\delaunay_n17.graph (4969106, 2018-12-21)
graphs\delaunay_n18.graph (10604804, 2018-12-21)
graphs\facebook_combined.txt (854362, 2018-12-21)
graphs\tiny (0, 2018-12-21)
graphs\tiny\test1.graph (45, 2018-12-21)
graphs\tiny\test2.graph (61, 2018-12-21)
graphs\tiny\test3.graph (223, 2018-12-21)
graphs\tiny\test4.graph (113, 2018-12-21)
graphs\tiny\test5.graph (42, 2018-12-21)
graphs\tiny\test6.graph (27, 2018-12-21)
kernels.cu (12339, 2018-12-21)
kernels.cuh (838, 2018-12-21)
main.cu (1534, 2018-12-21)
makefile (1369, 2018-12-21)
parse.cpp (9798, 2018-12-21)
parse.h (904, 2018-12-21)
sequential.cpp (1256, 2018-12-21)
sequential.h (209, 2018-12-21)
util.cu (3774, 2018-12-21)
util.cuh (1429, 2018-12-21)
hybrid_BC
=========
Hybrid methods for Parallel Betweenness Centrality on the GPU
The only dependency of this code (other than CUDA itself, of course) is Boost. A few small input examples can be found in the graphs directory. Both DIMACS and SNAP graph formats are accepted (assuming unweighted, undirected graphs). DIMACS files are assumed to end in the ".graph" extension whereas SNAP (edgelist) files are assumed to end in either ".txt" or ".edge" extensions.
Example command line to run and time the algorithm:
$ ./bc -i ./graphs/breast_coexpress.txt
To do the same and compare to sequential execution on the CPU:
$ ./bc -i ./graphs/breast_coexpress.txt -v
To print BC scores to a file:
$ ./bc -i ./graphs/breast_coexpress.txt --printscores=output.txt
If you use the algorithms in this repository for your own research, please cite our SC14 paper: https://dl.acm.org/citation.cfm?id=2683656
More information on the algorithms used in this repository can be found in the paper itself: http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.728.2926&rep=rep1&type=pdf
Additionally, this work was featured as a research highlight in the August issue of Communications of the ACM! https://cacm.acm.org/magazines/2018/8/229768-accelerating-gpu-betweenness-centrality/fulltext
Adam McLaughlin
Adam27X@gatech.edu
近期下载者:
相关文件:
收藏者: