Getting Started

Getting Started#

The simplest way to get started is by downloading one of the MultiMesh for JAX containers. There is a full development container which includes all source files and enables rebuilding each component.

Docker Images#

The easiest way to get started is by using the prebuilt container images:

docker pull ghcr.io/nv-legate/multimesh-jax:v0.2

Images can also be built using MultiMesh for Jax workflows.

Requirements#

The prebuilt container is built from a CUDA 12.8 base image on Ubuntu 22. For system compatibility, refer to the CUDA toolkit documentation.

Running Jupyter tutorials with Docker#

The recommended way to run the examples is through Docker. Note: prebuilt containers are not yet available for open-source. To launch a Jupyter notebook in the container for running on CPU that can be loaded in a local browser:

docker run \
  -w /opt/workspace/multimesh-jax/docs/notebooks \
  -p 8675:8675 \
  ghcr.io/nv-legate/multimesh-jax:v0.2 \
  jupyter notebook --allow-root --ip 0.0.0.0 --port=8675

The notebook will then be available at the URL shown, which is usually http://127.0.0.1:8675/... or http://localhost:8675/.... If GPUs are available, then docker can be launched as:

docker run \
  -w /opt/workspace/multimesh-jax/docs/notebooks \
  -p 8675:8675 \
  --gpus <N> \ 
  ghcr.io/nv-legate/multimesh-jax:v0.2 \
  jupyter notebook --allow-root --ip 0.0.0.0 --port=8675

where <N> is the number of GPUs.

Building Images#

Instructions for building MultiMesh for JAX can be found in the build monorepo that handles all of the components required for building the plugin.

Contents