.. _neuron-appnotes-index:
.. _neuron-appnotes:

.. meta::
   :noindex:
   :nofollow:
   :description: Archived AWS Neuron SDK application notes for support announcements, performance optimization, migration guides, and framework-specific implementations.
   :date-modified: 2026-07-21

Neuron application notes
========================

.. warning::

   This content is archived and no longer actively maintained. These
   application notes are provided for reference only and may describe
   unsupported features, deprecated APIs, or older SDK releases. For current
   documentation, see the `AWS Neuron SDK documentation
   <https://awsdocs-neuron.readthedocs-hosted.com/en/latest/>`_.

.. toctree:: 
   :maxdepth: 2
   :hidden:

   Neuron Runtime Library <neuron1x/introducing-libnrt>
   Mixed precision (neuron-cc) <neuron-cc/mixed-precision>
   Mixed precision (neuronx-cc) <neuronx-cc/neuronx-cc-training-mixed-precision>
   Introducing NxD Inference <neuronx-distributed/introducing-nxd-inference>
   Introducing NxD Training <neuronx-distributed/introducing-nxdt-training>
   Performance <perf/neuron-cc/performance-tuning>
   Parallel execution <perf/neuron-cc/parallel-ncgs>
   PyTorch for Neuron <torch-neuron/index>
   PyTorch for NeuronX <torch-neuronx/index>

Application notes provide specific documentation for support announcements, migration guides, performance optimization techniques, and framework-specific implementations for AWS Neuron SDK components.


Framework integration
---------------------

.. grid:: 1 1 2 2
   :gutter: 2

   .. grid-item-card::
      :link: torch-neuron-r-cnn-app-note
      :link-type: ref

      **PyTorch Neuron (Inf1)**
      ^^^
      R-CNN implementation and optimization techniques for PyTorch on ``Inf1``

   .. grid-item-card::
      :link: torch-neuronx-graph-partitioner-app-note
      :link-type: ref

      **PyTorch NeuronX Graph Partitioner**
      ^^^
      Advanced graph partitioning strategies for distributed training and inference

   .. grid-item-card::
      :link: torch-neuronx-dataparallel-app-note
      :link-type: ref

      **Data Parallel Inference on Torch NeuronX**
      ^^^
      Guide to using ``torch.neuronx.DataParallel`` for scalable inference on ``Inf1``

   .. grid-item-card::
      :link: torch-neuron-dataparallel-app-note
      :link-type: ref

      **Data Parallel Inference on Torch Neuron**
      ^^^
      Guide to using ``torch.neuron.DataParallel`` for scalable inference on ``Inf1``

   .. grid-item-card::
      :link: migration_from_xla_downcast_bf16
      :link-type: ref

      **Migrate from XLA_USE_BF16/XLA_DOWNCAST_BF16**
      ^^^
      Guide to migrating from deprecated XLA environment variables to recommended PyTorch mixed-precision options on NeuronX

   .. grid-item-card::
      :link: introduce-pytorch-2-9
      :link-type: ref

      **PyTorch 2.9 Support**
      ^^^
      New features and migration guide for PyTorch 2.9 on Neuron