|Project Name||Stars||Downloads||Repos Using This||Packages Using This||Most Recent Commit||Total Releases||Latest Release||Open Issues||License||Language|
|Easy3d||993||15 days ago||8||gpl-3.0||C++|
|A lightweight, easy-to-use, and efficient C++ library for processing and rendering 3D data|
|Quickqanava||939||12 days ago||April 16, 2018||31||other||C++|
|:link: C++14 network/graph visualization library / Qt node editor.|
|Metis||323||2 months ago||28||other||C|
|METIS - Serial Graph Partitioning and Fill-reducing Matrix Ordering|
|Libfirm||318||2 years ago||1||lgpl-2.1||C|
|graph based intermediate representation and backend for optimising compilers|
|Gbtl||114||a year ago||5||other||C++|
|GraphBLAS Template Library (GBTL): C++ graph algorithms and primitives using semiring algebra as defined at graphblas.org|
|Amdmigraphx||113||an hour ago||164||mit||C++|
|AMD's graph optimization engine.|
|Compareintegermaps||102||4 years ago||1||mit||C|
|Generates benchmark data for two different data structures, then renders some graphs.|
|Nnview||83||4 days ago||2||mit||C|
|A neural network visualizer|
|Gatb Core||57||7 months ago||12||C++|
|Core library of the Genome Analysis Toolbox with de-Bruijn graph|
|Samp Plugin Profiler||37||a year ago||9||bsd-2-clause||C++|
|Performance profiler plugin for SA-MP server|
Data analytics is a growing application of high-performance computing. Many advanced data analytics problems can be couched as graph problems. In turn, many of the common graph problems today can be couched as sparse linear algebra. This is the motivation for nvGraph, which harnesses the power of GPUs for linear algebra to handle large graph analytics.
This repository contains the legacy version of nvGraph as it was in the NVIDIA CUDA Toolkit. The aim is to provide a way for nvGraph users to continue using nvGraph after the CUDA Toolkit stops releasing it. While we still accept bug reports, we do not actively develop this product. If you find and can reproduce bugs in nvGRAPH, please report issues on GitHub.
Recently, NVIDIA started developing cuGraph a collection of graph analytics that process data found in GPU Dataframes as part of RAPIDS. Most nvGraph algorithms are now part of cuGraph too. In addition, cuGraph aims to provide a NetworkX-like API that will be familiar to data scientists, so they can now build GPU-accelerated workflows more easily. For more project details, see rapids.ai.
You can obtain CUDA from https://developer.nvidia.com/cuda-downloads. Compiler requirements:
It is easy to install nvGraph from source. As a convenience, a
build.sh script is provided. Run the script as shown below to download the source code, build and install the library. Note that the library will be installed to the location set in
export CUDA_ROOT=/usr/local/cuda). These instructions were tested on Ubuntu 18.04.
git clone https://github.com/rapidsai/nvgraph.git cd nvgraph export CUDA_ROOT=/usr/local/cuda ./build.sh # build the nvGraph library and install it to $CUDA_ROOT (you may need to add the sudo prefix)
The following instructions are for developers and contributors to nvGraph development. These instructions were tested on Linux Ubuntu 18.04. Use these instructions to build nvGraph from source and contribute to its development. Other operating systems may be compatible, but are not currently tested.
The nvGraph package is a C/C++ CUDA library. It needs to be installed in order for nvGraph to operate correctly.
The following instructions are tested on Linux systems.
To install nvGraph from source, ensure the dependencies are met and follow the steps below:
# Set the localtion to nvGraph in an environment variable NVGRAPH_HOME export NVGRAPH_HOME=$(pwd)/nvgraph # Download the nvGraph repo git clone https://github.com/rapidsai/nvgraph.git $NVGRAPH_HOME # Next load all the submodules cd $NVGRAPH_HOME git submodule update --init --recursive
libnvgraph_rapids.so. CMake depends on the
nvccexecutable being on your path or defined in
This project uses cmake for building the C/C++ library. To configure cmake, run:
cd $NVGRAPH_HOME cd cpp # enter nvgraph's cpp directory mkdir build # create build directory cd build # enter the build directory cmake .. -DCMAKE_INSTALL_PREFIX=$CONDA_PREFIX # now build the code make -j # "-j" starts multiple threads make install # install the libraries
The default installation locations are
# Run the tests cd $NVGRAPH_HOME cd cpp/build gtests/NVGRAPH_TEST # this is an executable file
These tests verify that the library was properly built and that the graph structure works as expected. We currently do not maintain the algorithm test suite. Most graph analytics features are now developed and tested in cuGraph.
The C API documentation can be found in the CUDA Toolkit Documentation.