PyTorch Geometric Container Early Access

PyG (PyTorch Geometric) is a library built upon PyTorch to easily write and train Graph Neural Networks (GNNs) for a wide range of applications related to structured data.

Image alt text

To enable developers to quickly take advantage of GNNs, we’ve partnered with the PyG team to provide a containerized solution that includes the latest PyG, PyTorch, NVIDIA RAPIDS™ (cudf, xgboost, rmm, cuml, and cugraph), and a set of tested dependencies.

It will be available in private early access Q4 2022.



Use the link below to sign up for the waitlist.


Join now

RAPIDs GPU Optimized Preprocessing

Go from hours to minutes. With NVIDIA RAPIDS™ integration, cuDF accelerates pandas queries so that you can run ETL with GPU-optimized code.

Read Guide

Multi GPU Support

Includes data parallel, distributed batching, and distributed sampling.

Read Guide

Tested Dependencies

NVIDIA-optimized PyG container is performance-tuned and tested for NVIDIA GPUs. This eliminates the need to manage packages and dependencies or build PyG from source. Containerized PyG, with all dependencies included, provides an easy place to start developing fraud detection, recommendation systems, and drug discovery applications.

Watch Session

NVIDIA’s Turnkey, End-to-End GPU Accelerated GNN Pipelines

NVIDIA provides a flexible, easy to use, general API that allows for building end to end workflows, addressing the top challenges with using GNNs today.

Watch Session