Graph Matching Gunrock

https://github.com/gunrock/docs/raw/master/docs/_media/logo.png

Gunrock GPU Graph Analytics

Note

This version of Gunrock is deprecated. Please use the main or develop branch for new features. Some algorithms from 1.x.x still need to be ported; for those, use the master or dev branch.

Gunrock is a CUDA library for graph processing designed for the GPU. It utilizes a high-level, bulk-synchronous, data-centric abstraction focusing on operations on vertex or edge frontiers. Gunrock achieves a balance between performance and expressiveness by coupling high-performance GPU computing primitives and optimization strategies. It also offers a high-level programming model that allows programmers to quickly develop new graph primitives that scale from one to many GPUs on a node with minimal GPU programming knowledge. For more details, see Gunrock’s Overview.

Quick Start Guide

Before building Gunrock, make sure you have CUDA Toolkit 10.2 or higher installed on your Linux system. You can also build Gunrock using Docker images provided in the docker subdirectory. For a complete build guide, see Building Gunrock.

git clone --recursive https://github.com/gunrock/gunrock/
cd gunrock
mkdir build && cd build
cmake .. && make -j$(nproc)
make test

Getting Started with Gunrock