Getting Started#

Installing pre-built conda packages#

numba-dpex along with its dependencies can be installed using conda. It is recommended to use conda packages from the channel to get the latest production releases.

conda create -n numba-dpex-env                                             \
    numba-dpex dpnp dpctl dpcpp-llvm-spirv                                 \
    -c intel -c conda-forge

To try out the bleeding edge, the latest packages built from tip of the main source trunk can be installed from the dppy/label/dev conda channel.

conda create -n numba-dpex-env                                             \
    numba-dpex dpnp dpctl dpcpp-llvm-spirv                                 \
    -c dppy/label/dev -c intel -c conda-forge

Building from source#

numba-dpex can be built from source using either conda-build or setuptools (with scikit-build backend).

Steps to build using conda-build:

  1. Ensure conda-build is installed in the base conda environment or create a new conda environment with conda-build installed.

conda create -n build-env conda-build
conda activate build-env
  1. Build using the vendored conda recipe

conda build conda-recipe -c intel -c conda-forge
  1. Install the conda package

conda install -c local numba-dpex

Steps to build using

As before, a conda environment with all necessary dependencies is the suggested first step.

# Create a conda environment that hass needed dependencies installed
conda create -n numba-dpex-env                                             \
    scikit-build cmake dpctl dpnp numba dpcpp-llvm-spirv llvmdev pytest    \
    -c intel -c conda-forge
# Activate the environment
conda activate numba-dpex-env
# Clone the numba-dpex repository
git clone
cd numba-dpex
python develop

Building inside Docker#

A Dockerfile is provided on the GitHub repository to build numba-dpex as well as its direct dependencies: dpctl and dpnp. Users can either use one of the pre-built images on the numba-dpex GitHub page or use the bundled Dockerfile to build numba-dpex from source. Using the Dockerfile also ensures that all device drivers and runtime libraries are pre-installed.


Numba dpex ships with multistage Dockerfile, which means there are different targets available for build. The most useful ones:

  • runtime

  • runtime-gpu

  • numba-dpex-builder-runtime

  • numba-dpex-builder-runtime-gpu

To build docker image

docker build --target runtime -t numba-dpex:runtime ./

To run docker image

docker run -it --rm numba-dpex:runtime


When trying to build a docker image with Intel GPU support, the Dockerfile will attempt to use the GitHub API to get the latest Intel GPU drivers. Users may run into an issue related to Github API call limits. The issue can be bypassed by providing valid GitHub credentials using the GITHUB_USER and GITHUB_PASSWORD build args to increase the call limit. A GitHub access token can also be used instead of the password.


When building the docker image behind a firewall the proxy server settings should be provided using the http_proxy and https_proxy build args. These build args must be specified in lowercase.

The bundled Dockerfile supports different python versions that can be specified via the PYTHON_VERSION build arg. By default, the docker image is based on the official python image based on slim debian. The requested python version must be from the available python docker images.

The BASE_IMAGE build arg can be used to build the docker image from a custom image. Note that as the Dockerfile is based on debian any custom base image should be debian-based, like debian or ubuntu.

The list of other build args are as follows. Please refer the Dockerfile to see currently all available build args.


  • CR_TAG


  • CM_TAG

  • L0_TAG














  • http_proxy

  • https_proxy




Using the pre-built images#

There are several pre-built docker images available:

  • runtime package that provides a pre-built environment with numba-dpex

    already installed. It is ideal to quickly setup and try numba-dpex.<numba_dpex_version>-py<python_version>[-gpu]
  • builder package that has all required dependencies pre-installed and is

    ideal for building numba-dpex from source.<numba_dpex_version>-py<python_version>[-gpu]
  • stages package primarily meant for creating a new docker image that is

    built on top of one of the pre-built images.

After setting up the docker image, to run numba-dpex the following snippet can be used.

docker run --rm -it bash

It is advisable to verify the SYCL runtime and driver installation within the container by either running,



python -m dpctl -f


To enable GPU device, the device argument should be used and one of the *-gpu images should be used.

For passing GPU into container on linux use arguments --device=/dev/dri. However if you are using WSL you need to pass --device=/dev/dxg -v /usr/lib/wsl:/usr/lib/wsl instead.

For example, to run numba-dpex with GPU support on WSL:

docker run --rm -it \
--device=/dev/dxg -v /usr/lib/wsl:/usr/lib/wsl \


numba-dpex uses pytest for unit testing and the following example shows a way to run the unit tests.

python -m pytest --pyargs numba_dpex.tests


A set of examples on how to use numba-dpex can be found in numba_dpex/examples.