Skip to content

Latest commit

 

History

History
50 lines (29 loc) · 3.02 KB

Deployment_Guide.md

File metadata and controls

50 lines (29 loc) · 3.02 KB
Raw

NVIDIA Holoscan Deployment Guide

This Bootcamp focuses on building an end-to-end AI-enabled streaming pipeline using the Holoscan SDK, handling sensor I/O, applying a trained AI model to a real time sensor stream, and building GPU accelerated applications. We will also discuss techniques to measure application performance and transition from prototype to production.

Deploying the materials

Prerequisites

To run this tutorial, you will need a machine with NVIDIA GPU.

  • Install the latest Docker or Singularity.

  • The base containers required for the lab may require users to create an NGC account and generate an API key.

The material has been tested to be working with NVIDIA A100 GPU, please contact us if you require assistance in deploying the content.

Tested environment

These materials was tested with both Docker and Singularity on an NVIDIA A100 GPU in an x86-64 platform installed with a driver version of 535.104.05.

Deploying with container

These materials can be deployed with either Docker or Singularity container, refer to the respective sections for the instructions.

Docker Container

  • To build the docker container for NVIDIA Holoscan bootcamp, follow the below steps:

    1. sudo docker build . -t holoscan:openhackathons -f Dockerfile --network host
    2. sudo docker run -it --rm --ipc=host --runtime=nvidia --gpus all --ulimit memlock=-1 --ulimit stack=67108864 -p 8888:8888 holoscan:openhackathons
    3. Now, open the jupyter lab in browser: http://localhost:8888, and start working on the lab by clicking on the Python-Holoscan-Tutorial.ipynb notebook

Docker Container using rootless Docker

  • To build the container using rootless Docker on a cluster (e.g. Curiosity) use Dockerfile_cluster:

    1. sudo docker build . -t holoscan:openhackathons -f Dockerfile_cluster --network host
    2. sudo docker run -it --rm --ipc=host --runtime=nvidia --gpus all --ulimit memlock=-1 --ulimit stack=67108864 -p 8888:8888 holoscan:openhackathons
    3. Now, open the jupyter lab in browser: http://localhost:8888, and start working on the lab by clicking on the Python-Holoscan-Tutorial.ipynb notebook

Singularity Container

  • To build the singularity container for NVIDIA Holoscan bootcamp, follow the below steps:

    1. singularity build --fakeroot --sandbox holoscan:openhackathons singularity
    2. singularity run --writable --nv holoscan:openhackathons jupyter-lab --no-browser --allow-root --ip=0.0.0.0 --port=8888 --NotebookApp.token="" --notebook-dir=/workspace/python
    3. Now, open the jupyter lab in browser: http://localhost:8888, and start working on the lab by clicking on the Python-Holoscan-Tutorial.ipynb notebook

Known issues

  • Please go through the list of exisiting bugs/issues or file a new issue at Github.