TC-GNN (Running Sparse GNN on Dense Tensor Core on Ampere GPU)

• Cite this project and paper.

@inproceedings{TC-GNN,
  title={TC-GNN: Accelerating Sparse Graph Neural Network Computation Via Dense Tensor Core on GPUs},
  author={Yuke Wang and Boyuan Feng and Zheng Wang and Guyue Huang and Yufei Ding},
  booktitle={USENIX Annual Technical Conference},
  year={2023}
}

• Clone this project.

git clone [email protected]:YukeWang96/TCGNN-Pytorch.git

• OS & Compiler:

• Ubuntu 16.04+
• gcc >= 7.5
• cmake >= 3.14
• CUDA >= 11.0 and nvcc >= 11.0

Files and Directories.

• config.py: the configuration file for the shape of a TC block.
• bench.py: the benchmark file for invoking main_tcgnn.py for various datasets and models.
• main_tcgnn.py: the main entry for running TC-GNN.
• count_TC_blocks.py: counting the total number of TC blocks without sparse-graph translation.
• proc_prof.py: get the detailed GPU kernel metrics from the ncu csv output.
• TCGNN_conv/: the directory for core TC-GNN implementations, including TCGNN_kernel.cu and TCGNN.cpp.

Environment Setup.

[Method-1] Install via Docker (Recommended).

• Go to Docker/
• Run ./build.sh
• Run ./launch.sh

[Method-2] Install via Conda.

• Install conda on system Toturial.
• Create a conda environment:

conda create -n env_name python=3.6

• Install Pytorch:

conda install pytorch torchvision torchaudio cudatoolkit=11.1 -c pytorch -c conda-forge

or using pip [Note that make sure the pip you use is the pip from current conda environment. You can check this by which pip]

pip install torch==1.8.0+cu111 torchvision==0.9.0+cu111 torchaudio==0.8.0 -f https://download.pytorch.org/whl/torch_stable.html

• Install Deep Graph Library (DGL).

conda install -c dglteam dgl-cuda11.0
pip install torch requests tqdm

• Install Pytorch-Geometric (PyG).

pip install torch-scatter -f https://pytorch-geometric.com/whl/torch-1.8.0+cu111.html
pip install torch-sparse -f https://pytorch-geometric.com/whl/torch-1.8.0+cu111.html
pip install torch-cluster -f https://pytorch-geometric.com/whl/torch-1.8.0+cu111.html
pip install torch-spline-conv -f https://pytorch-geometric.com/whl/torch-1.8.0+cu111.html
pip install torch-geometric

Install TC-GNN.

Go to TCGNN_conv/, then run

./build.sh

to install the TCGNN_conv modules with Pytorch binding. Note that this step is required for both Docker and Conda setup.

Download graph datasets.

Get the preprocessed datasets in .npy at here, then run

wget https://storage.googleapis.com/graph_dataset/tcgnn-ae-graphs.tar.gz
tar -zxvf tcgnn-ae-graphs.tar.gz

Running PyG baseline.

• Go to pyg_baseline/ directory;
• Pass the --model parameter in pyg_main.py with gcn and gin to profile the example GCN and GIN model, respectively;
• ./0_bench.py| tee run_pyg.log to run the script and the report 10 epoch runtime for all evaluated datasets.
• ./1_log2csv.py to convert the run_pyg.log to run_pyg.csv for ease of analysis.

Running DGL baseline.

• Go to dgl_baseline/ directory
• Pass the --model parameter in dgl_main.py with gcn and gin to profile the example GCN and GIN model, respectively;
• ./0_bench.py| tee run_dgl.log to run the script and the report 10 epoch runtime for all evaluated datasets.
• ./1_log2csv.py to convert the run_dgl.log to run_dgl.csv for ease of visualization.

Running TC-GNN.

• Under the current project directory
• ./0_bench.py| tee run_TCGNN.log to run the script and the report 10 epoch runtime for all evaluated datasets.
• ./1_log2csv.py to convert the run_TCGNN.log to run_TCGNN.csv for ease of analysis.