DNN (Deep Neural Network)
=========================

.. toctree::
   :maxdepth: 1
   :caption: Contents:

.. warning:: Only CPU DNN engines are supported on Discoverer HPC nodes. GPU DNN engines are not yet supported!

Intel® oneDNN
-------------

`Intel® oneAPI Deep Neural Network Library`_ or oneDNN is included in the installation of Intel® oneAPI supported on Discoverer HPC.

Originally, oneDNN should work as expected on Intel CPU families, but it also performs well on the `AMD EPYC 7H12 64-Core Processors`_ installed on Discoverer HPC compute nodes, unless `AVX-512`_ support is essential for running the code.

Loading the environment for supporting the use of oneAPI is a two-stage process. First, the basic environment modules have to be loaded:

.. code:: bash

   module load intel
   module load dnnl/latest

Then load the module providing access to the thread library adopted by the code:

* GNU OpenMP library

.. code:: bash

   module load dnnl-cpu-gomp/latest

* Intel OpenMP library

.. code:: bash

   module load dnnl-cpu-iomp/latest

* `Intel® TBB library`_

.. code:: bash

   module load dnnl-cpu-tbb/latest

AMD ZenDNN
----------

`AMD ZenDNN`_ is built (against AMD AOCC) and supported by Discoverer HPC team. It is available in the software repository.

To gain access to the installation, load the corresponding environment module:

.. code:: bash

  module load amd/zendnn/3/latest

MagmaDNN
--------

`MagmaDNN`_ is built (against a variety of compilers and BLAS/LAPACK libraries) and supported by the Discoverer HPC team. It is available in the software repository.

To gain access to the installation that corresponds to the employed compiler, load the specific environment module:

Intel® oneAPI
.............

.. code:: bash

   module load magmadnn/1/latest-intel

NVIDIA HPC SDK
..............

.. code:: bash

   module load magmadnn/1/latest-nvidia

AMD AOCC
........

.. code:: bash

   module load magmadnn/1/latest-aocc

GNU GCC
.......

.. code:: bash

   module load magmadnn/1/latest-gcc

User-supported DNN
------------------

Users are welcome to install (using Conda, for instance) other DNN types or builds, but they will not be supported by the Discoverer HPC team.

Getting help
------------

See :doc:`help`

.. _`Intel® oneAPI Deep Neural Network Library`: https://www.intel.com/content/www/us/en/developer/tools/oneapi/onednn.html
.. _`AMD EPYC 7H12 64-Core Processors`: https://www.amd.com/en/products/cpu/amd-epyc-7h12
.. _`AVX-512`: https://www.intel.com/content/www/us/en/architecture-and-technology/avx-512-animation.html
.. _`Intel® TBB library`: https://www.intel.com/content/www/us/en/developer/tools/oneapi/onetbb.html
.. _`AMD ZenDNN`: https://github.com/amd/ZenDNN
.. _`MagmaDNN`: https://github.com/MagmaDNN/magmadnn