Skip to content

Bottleneck Analysis of Dynamic Graph Neural Networks Performance on Hardware

Notifications You must be signed in to change notification settings

sharc-lab/DGNN_analysis

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

69 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Dynamic Graph Neural Networks on Hardware: Bottleneck Analysis (IISWC 2022)

This repository contains the code for Dynamic Graph Neural Networks on Hardware: Bottleneck Analysis (IISWC 2022).


NVIDIA Nsight Systems Installation

Please install the profiling tool using this link: https://developer.nvidia.com/nsight-systems

Directly run the downloaded installer to install this tool on your machine.

Pytorch Profiler Installation

1. Install PyTorch and Torchvision using the following command:

pip install torch torchvision

2. Install PyTorch Profiler TensorBoard Plugin for visualization:

pip install torch_tb_profiler

3. Open the generated trace file using this command:

tensorboard --logdir=./log

log is a folder containing trace files. You can set it to any path containing the generated trace files.

Profiling results:

Once you profile the models using Pytorch Profiler, you will get results as follows:

image

If you choose to use NVIDIA Nsight Systems, you will get similiar results below:

image

About

Bottleneck Analysis of Dynamic Graph Neural Networks Performance on Hardware

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Python 99.8%
  • Shell 0.2%