NVIDIA NGC

NVIDIA GPU Cloud (NGC) offers pre-built containers

They have both general-purpose and domain-specific offerings for machine learning and deep learning workloads on NVIDIA GPUs.

Domino enhances NGC containers for use in Domino and for general data science work.

Add Domino compatibility

Domino automatically manages code and data versioning as part of the container lifecycle and as a result we require specific additional software in the container.

Add Data Source drivers

Domino adds drivers for Data Sources such as Snowflake, Oracle, and Microsoft SQL to make it easy to connect to a Data Source without needing to manually add drivers.

Add Workspaces

When an NGC container doesn’t include an interactive notebook, Domino adds Jupyter to the image so you can interactively develop in the image. When there’s a notebook already included, Domino configures the existing notebook to work.

Domino builds the latest versions of NGC containers with a preference for Ubuntu base images. If there’s an Environment that you’re looking for which is not available, request it from ngc-request@dominodatalab.com.

Add an NGC image to Domino

  1. Choose the NGC container from the list of available containers.

  2. Create a new Compute Environment in Domino.

  3. Use the link from the container you selected as the base compute Environment (for example, quay.io/domino/ngc-pytorch:20.12-py3)

  4. Enter the following in the Workspace Definition area:

jupyter:
 title: "Jupyter (Python, R, Julia)"
 iconUrl: "/assets/images/workspace-logos/Jupyter.svg"
 start: [ "/opt/domino/workspaces/jupyter/start" ]
 httpProxy:
   port: 8888
   rewrite: false
   internalPath: "/{{ownerUsername}}/{{projectName}}/{{sessionPathComponent}}/{{runId}}/{{   requireSubdomain: false
jupyterlab:
 title: "JupyterLab"
 iconUrl: "/assets/images/workspace-logos/jupyterlab.svg"
 start: [  "/opt/domino/workspaces/jupyterlab/start" ]
 httpProxy:
   internalPath: "/{{ownerUsername}}/{{projectName}}/{{sessionPathComponent}}/{{runId}}/{{   port: 8888
   rewrite: false
   requireSubdomain: false

Available containers

RAPIDS

https://ngc.nvidia.com/catalog/containers/nvidia:rapidsai:rapidsai

Domino registry path:

quay.io/domino/ngc-rapids:0.18-cuda11.0-runtime-ubuntu20.04-py3.8

Versions:

RAPIDS 0.18, Ubuntu 18.04, CUDA 11

Base Image:

docker pull nvcr.io/nvidia/rapidsai/rapidsai:0.18-cuda11.0-base-ubuntu18.04

Notes:

N/A

PyTorch

https://ngc.nvidia.com/catalog/containers/nvidia:pytorch

Domino registry path:

quay.io/domino/ngc-pytorch:20.12-py3

Versions:

Ubuntu 18.04, CUDA 11

Base Image:

docker pull nvcr.io/nvidia/pytorch:21.02-py3

Notes:

N/A

TensorFlow

https://ngc.nvidia.com/catalog/containers/nvidia:tensorflow

Domino registry path:

quay.io/domino/ngc-tensorflow:20.12-tf1-py3

Versions:

TensorFlow v21.03, Ubuntu 18.04, CUDA 11

Base Image:

docker pull nvcr.io/nvidia/tensorflow:21.03-tf1-py3

Notes:

N/A

Clara Train

https://ngc.nvidia.com/catalog/containers/nvidia:clara-train-sdk

Domino registry path:

quay.io/domino/ngc-clara-train:v3.1.01

Versions:

Clara v3.1.01, Ubuntu 18.04, CUDA 11

Base Image:

docker pull nvcr.io/nvidia/clara-train-sdk:v3.1.01

Notes:

N/A

MXNet

https://ngc.nvidia.com/catalog/containers/nvidia:mxnet

Domino registry path:

quay.io/domino/ngc-mxnet:20.12-py3

Versions:

MXNet v21.03, Ubuntu 18.04, CUDA 11

Base Image:

docker pull nvcr.io/nvidia/mxnet:21.03-py3

Notes:

N/A