instructions.md

This directory contains Jupyter notebooks that can be run on a local installation of CUDA-Q. The requirements.txt and Dockerfile are included here. Please refer to the Quick Start Guide for instructions on how to install CUDA-Q on your system.

Most of the material in these notebooks can be run without a GPU. However, the portions of the notebook that use MPI will require a GPU to execute.

If you don't have CUDA-Q installed on your system, you can run the notebooks in Google Colab.

Building the container for local execution

The following command will build the CUDA Quantum Academic container. To customize this container, make edits to the included Dockerfile.

# Login to NVIDIA GPU Cloud for access to CUDA-Q base container
docker login nvcr.io
# Follow the login instructions at ngc.nvidia.com
# Next, build the container locally
docker build -t cuda-quantum-academic:latest

To run the container, use the following command. By default Jupyter-lab will use port 8888 and docker will expose this port. If you wish to use a different port, see the directions below.

docker run cuda-quantum-academic:latest

You can now open a web browser to http://localhost:8888/lab to access the labs.

Changing the port

You can either change the port that will be used by jupyter-lab at build time (more permanent) or at runtime (more dynamic).

Build time:

docker build --build-arg JUPYTER_LAB_PORT=8000 -t cuda-quantum-academic:latest
docker run cuda-quantum-academic:latest

Run time:

docker run -p 8000:8888 cuda-quantum-academic:latest

Running the notebooks in Google Colab

Simply click on the icon at the top of each notebook in github to open it up in Google Colab. In each notebook there instructions for running CoLab.