graph-bench

Benchmarks suite for performance study of various graph analysis frameworks for CPU/GPU computations.

Tools description

Name	Brief	Platform	Technology	Source Page
Spla	Generalized linear sparse linear algebra for multi-GPU computation	GPU	OpenCL	link
GraphBLAST	High-performance linear algebra-based graph primitives on GPUs	GPU	CUDA	link
Gunrock	High-performance graph primitives on GPUs	GPU	CUDA	link
LaGraph	Collection of graph algorithms for SuiteSparse:GraphBLAS libray	CPU	OpenMP	link

Dataset description

Name	Vertices	Edges	Avg Deg	Sd Deg	Min Deg	Max Deg	Link
coAuthorsCiteseer	227.3K	1.6M	7.2	10.6	0.0	1372.0	link
mycielskian19	393.2K	903.2M	2296.4	3950.8	0.0	196606.0	link
coPapersDBLP	540.5K	30.5M	56.4	66.2	0.0	3299.0	link
amazon-2008	735.3K	7.0M	9.6	7.6	0.0	1077.0	link
hollywood-2009	1.1M	112.8M	98.9	271.9	0.0	11467.0	link
belgium_osm	1.4M	3.1M	2.2	0.5	0.0	10.0	link
roadNet-CA	2.0M	5.5M	2.8	1.0	0.0	12.0	link
com-Orkut	3.1M	234.4M	76.3	154.8	0.0	33313.0	link
cit-Patents	3.8M	33.0M	8.8	10.5	0.0	793.0	link
rgg_n_2_22_s0	4.2M	60.7M	14.5	3.8	0.0	36.0	link
soc-LiveJournal	4.8M	85.7M	17.7	52.0	0.0	20333.0	link
indochina-2004	7.4M	302.0M	40.7	329.6	0.0	256425.0	link
rgg_n_2_23_s0	8.4M	127.0M	15.1	3.9	0.0	40.0	link
road_central	14.1M	33.9M	2.4	0.9	0.0	8.0	link
twitter7	41.7M	2405.0M	57.7	401.5	0.0	2997490.0	link

Instructions

1. How to get source code

Download benchmark repository source code.

git clone https://github.com/EgorOrachyov/graph-bench.git

Within repo folder init git submodule to get all source code of tools.

git submodule update --init --recursive

2. How to build tools

2.1 Spla

Build bundled Spla library.

python3 scripts/build_spla.py

2.2 Gunrock

Build bundled Gunrock library.

python3 scripts/build_gunrock.py

2.3 GraphBLAST

Build bundled GraphBLAST library.

python3 scripts/build_graphblast.py

2.4 LaGraph

Build bundled SuiteSparse and LaGraph libraries.

python3 scripts/build_lagraph.py

3. How to download data

Download all graphs one by one archives and extract into dataset folder. Alternatively, download all graphs within single archive from Google Drive.

5. How to prepare data

After dataset unpack into dataset folder you have to run convert tool to prepare graphs.

python3 scripts/convert.py

This tool uses spla convert.exe to convert source .mtx files into undirected .mtx graphs.

6. How to run benchmarks

Run all algorithms & graphs & tools performance measurements.

python3 scripts/benchmark.py

Run particular tool for performance measurements. Use comma and no space to select multiple.

python3 scripts/benchmark.py --tool=[all, spla, lagraph, gunrock, graphblast]

python3 scripts/benchmark.py --tool=spla,lagraph,gunrock,graphblast

Run particular algorithm for performance measurements. Use comma and no space to select multiple.

python3 scripts/benchmark.py --algo=[all, bfs, sssp, pr, tc]

python3 scripts/benchmark.py --algo=bfs,sssp,pr,tc

Provide csv file name to save all stats of the benchmark.

python3 scripts/benchmark.py --csvall=my_results.csv

Provide csv file name to save per-tool detailed stats of the benchmark.

python3 scripts/benchmark.py --csvtool=my_results.csv

See help for more options.

python3 scripts/benchmark.py -h

License

This project licensed under MIT License. License text can be found in the license file.