Alveo Getting Started Guide¶
- A version of the Xilinx Runtime (XRT) above or equal
2.3installed in the system. Previous versions of XRT might still work, but are not explicitly supported. Moreover, the functionalities offered by the Embedded Runtime Library (ERT) will not work with versions of XRT below
- Any XRT-supported version of either RedHat/CentOS or Ubuntu as Operating System
- Python and PIP must be installed. The minimum Python version is
3.5.2, although the recommended minimum version is
pynq through PIP
pip install pynq
In case needed, please read the Extended Setup Instructions section for more detailed instructions.
Get the Introductory Examples¶
pynq-alveo-examples package using PIP
pip install pynq-alveo-examples
Once that is done, run the
get-notebooks command in your shell session
This will deliver all the available notebooks in a
pynq-notebooks folder in
your current working directory.
You can now move to the newly created
pynq-notebooks folder and run Jupyter
cd pynq-notebooks jupyter notebook
When retrieving the notebooks using the
command, overlays might be downloaded from the web and might be available
only for specific cards/shells. The
get-notebooks command has a few
optional parameters that can be used to customize the notebooks delivery.
pynq get-notebooks -h to see them. Refer to the
PYNQ Command Line Interface section for more detailed information.
Extended Setup Instructions¶
The first thing you will have to do before every session, is source the XRT setup script. To do so, open up a bash shell and type:
/opt/xilinx/xrt is the predefined install path for XRT and should
not be changed. Therefore, the setup script will always be located there.
In case you try to use PYNQ without sourcing XRT, you will get a warning asking if XRT was correctly sourced.
For instance, to install the latest
Anaconda distribution you can do
wget https://repo.anaconda.com/archive/Anaconda3-2019.10-Linux-x86_64.sh bash Anaconda3-2019.10-Linux-x86_64.sh
After you have installed it make sure conda is in your
PATH, and in case
is not just source the conda activation script
Using a Conda Environment¶
In case you want to use a conda environment instead of the base installation, follow these simple steps to get everything you need:
1. Save the content of this GIST as
pynq-envenvironment using the above configurationconda env create -f environment.yml
Activate the newly created environmentconda activate pynq-env
The provided environment.yml can also be useful to re-create an environment which is already tested and confirmed to be working, in case you are having issues.
By default, installing
pynq will not install
jupyter. In case you want
it, you can install it using PIP
pip install jupyter
Or install the
pynq-alveo-examples package as previously shown. This package
will install Jupyter as a dependency, alongside the other packages required to
run the included example notebooks.
When installing jupyter with a version of Python less than
you will have to make sure to have a compatible version of
installed. Therefore, in this case after installing
ipython with an appropriate version. The recommended is
7.9, and you can ensure this is the version installed by
pip install --upgrade ipython==7.9.