overhaul elements of index.rst and inst.rst #497

bconsolvo · 2025-10-10T14:19:56Z

Overhaul of index.rst and inst.rst for ease of use for developers

Keeping release notes with only actual release notes, not hardware configurations. Moved hardware configuration and model support to index (starting page), and adjusted other links elsewhere to point to this place.
Added HW support table to provide much more clarity on codenames and supported OS for 1.6.0 release
Added links for installation into table of prerequisites
Added code for setting environment variables in PowerShell (not obvious from documentation)
Added 5-step development workflow overview cycle to clarify a typical development cycle
Other small syntactical and structural changes

merging changes to highlight automatic BF16 conversion

Copilot

Pull Request Overview

This PR overhauls the documentation structure for index.rst and inst.rst to improve ease of use for developers. Key changes include reorganizing hardware configuration information, adding clearer installation instructions, and providing a structured development workflow overview.

Moved hardware configuration and model support information from release notes to the main index page
Enhanced installation instructions with specific links and PowerShell environment variable setup
Added a 5-step development workflow overview for typical Ryzen AI usage
Updated example titles and restructured documentation organization

Reviewed Changes

Copilot reviewed 7 out of 7 changed files in this pull request and generated 4 comments.

Show a summary per file

File	Description
docs/relnotes_backup.rst	Created as backup containing original release notes content with hardware configurations
docs/relnotes.rst	Removed hardware configuration sections to focus on actual release notes
docs/inst.rst	Enhanced with detailed prerequisites table, PowerShell setup code, and clearer installation steps
docs/index.rst	Added comprehensive hardware support tables and development workflow overview
docs/getstartex.rst	Updated tutorial title and added more detailed explanations
docs/examples.rst	Simplified structure and updated example titles
docs/conf.py	Added sphinx_copybutton extension for code block copying

_{Tip: Customize your code reviews with copilot-instructions.md. Create the file or learn how to get started.}

Copilot · 2025-10-15T03:57:23Z

docs/inst.rst

  - Specify the name for the conda environment (default: ``ryzen-ai-1.6.0``)

-The Ryzen AI Software packages are now installed in the conda environment created by the installer.
+The Ryzen AI Software packages should now installed in the conda environment created by the installer.


Missing word 'be' in sentence. Should read 'should now be installed'.

Suggested change

The Ryzen AI Software packages should now installed in the conda environment created by the installer.

The Ryzen AI Software packages should now be installed in the conda environment created by the installer.

Copilot · 2025-10-15T03:57:23Z

docs/inst.rst


- Download and Install the NPU driver version: 32.0.203.280 or newer using the following links: 
+- Under "Task Manager" in Windows, go to Performance -> NPU0 to check the driver version. 
+- If needed, download the NPU driver version: 32.0.203.280 or the latest 32.0.203.304 here:


The instruction mentions 'here:' but the actual download links are on the following lines. Consider rephrasing to 'download the NPU driver from one of the following links:' for better clarity.

Suggested change

- If needed, download the NPU driver version: 32.0.203.280 or the latest 32.0.203.304 here:

- If needed, download the NPU driver version: 32.0.203.280 or the latest 32.0.203.304 from one of the following links:

Copilot · 2025-10-15T03:57:24Z

docs/getstartex.rst

+*The CIFAR-10 dataset consists of 60,000 32x32 colour images in 10 classes, with 6,000 images per class. There are 50,000 training images and 10,000 test images.* You can learn more about the CIFAR-10 dataset here: https://www.cs.toronto.edu/~kriz/cifar.html. This dataset is used in the subsequent steps for quantization and inference. The script also exports the provided PyTorch model into ONNX format. The following snippet from the script shows how the ONNX model is exported:



[nitpick] The CIFAR-10 dataset description is formatted with asterisks instead of proper reStructuredText formatting. Consider using proper emphasis markup or a note directive for better presentation.

Suggested change

*The CIFAR-10 dataset consists of 60,000 32x32 colour images in 10 classes, with 6,000 images per class. There are 50,000 training images and 10,000 test images.* You can learn more about the CIFAR-10 dataset here: https://www.cs.toronto.edu/~kriz/cifar.html. This dataset is used in the subsequent steps for quantization and inference. The script also exports the provided PyTorch model into ONNX format. The following snippet from the script shows how the ONNX model is exported:

.. note::

The CIFAR-10 dataset consists of 60,000 32x32 colour images in 10 classes, with 6,000 images per class. There are 50,000 training images and 10,000 test images.

You can learn more about the CIFAR-10 dataset here: https://www.cs.toronto.edu/~kriz/cifar.html.

This dataset is used in the subsequent steps for quantization and inference. The script also exports the provided PyTorch model into ONNX format. The following snippet from the script shows how the ONNX model is exported:

Copilot · 2025-10-15T03:57:24Z

docs/getstartex.rst

 The C++ source files, CMake list files, and related artifacts are provided in the ``cpp/resnet_cifar/*`` folder. The source file ``cpp/resnet_cifar/resnet_cifar.cpp`` takes 10 images from the CIFAR-10 test set, converts them to .png format, preprocesses them, and performs model inference. The example has onnxruntime dependencies that are provided in ``%RYZEN_AI_INSTALLATION_PATH%/onnxruntime/*``.

 Run the following command to build the resnet example. Assign ``-DOpenCV_DIR`` to the OpenCV build directory.



Path uses backslashes which are Windows-specific. Consider using forward slashes for cross-platform compatibility or noting this is Windows-specific.

Suggested change

.. note::

The following command uses Windows-style backslashes and is intended for use in a Windows environment.

ThomasXilinx · 2025-10-15T16:42:47Z

docs/examples.rst

 NPU
 ~~~

- :doc:`Getting Started Tutorial for INT8 models <getstartex>` - Uses a custom ResNet model to demonstrate: 


Why remove "Getting Started" from the description?

ThomasXilinx · 2025-10-15T16:49:57Z

docs/index.rst

+     - 2025
+     - ☑️
+     - 
+   * - Ryzen Z2


I don't believe this device is supported.

ThomasXilinx · 2025-10-15T16:52:13Z

docs/index.rst

+****************
+Ryzen AI 1.6 Software runs on AMD processors outlined below. For a more detailed list of supported devices, refer to the `processor specifications <https://www.amd.com/en/products/specifications/processors.html>`_ page (scroll to the "AMD Ryzen™ AI" column toward the right side of the table, and select "Available" from the pull-down menu). Support for Linux is coming soon in Ryzen AI 1.6.1.
+
+.. list-table:: Supported Ryzen AI Processor Configurations


This table will only grow and will be hard to maintain as we add support for more platforms. It's redundant with the https://www.amd.com/en/products/specifications/processors.html page. And we risk creating inconsistencies. Case in point, see the comment about Z2 below.

I recommend we simply link to the official processor specification page.

ThomasXilinx · 2025-10-15T16:54:28Z

docs/index.rst

+     - GPU
+     - NPU
+     - Hybrid (NPU + iGPU)
+   * - Ryzen AI 300


We support LLMs on all STX and KRK platforms, not all Ryzen AI 300. The first column in this table is not needed and should be removed.

ThomasXilinx · 2025-10-15T16:56:31Z

docs/index.rst

+.. list-table::
+   :header-rows: 1
+
+   * - Model Type


Why mention CPU and GPU for LLMs and not for other models? BF16 models can run on CPU and GPU on PHX/HPT.
The way LLMs and CNN/NLPs are presented in inconsistent. It would be preferable to find a common way to presenting the information.

ThomasXilinx · 2025-10-15T16:59:24Z

docs/index.rst

 *************************

-The Ryzen AI development flow does not require any modifications to the existing model training processes and methods. The pre-trained model can be used as the starting point of the Ryzen AI flow.
+A typical Ryzen AI flow might look like the following:


This is accurate for CNNs, but not for BF16 NLPs (no quantization step) or LLMs (OGA flow).

ThomasXilinx · 2025-10-15T17:04:15Z

docs/inst.rst

+     - 2022 with `Desktop Development with C++` checked
+   * - `cmake <https://cmake.org/download/>`_
+     - >= 3.26
+   * - `Python (Miniforge preferred) <https://conda-forge.org/download/>`_


Should we really say "Miniforge preferred"?
Internally to AMD, we need to use Miniforge. But other companies may have different requirements.

ThomasXilinx · 2025-10-15T17:04:38Z

docs/inst.rst

- Miniforge: ensure that the following path is set in the System PATH variable: ``path\to\miniforge3\condabin`` or ``path\to\miniforge3\Scripts\`` or ``path\to\miniforge3\`` (The System PATH variable should be set in the *System Variables* section of the *Environment Variables* window).
+  $existingPath = [System.Environment]::GetEnvironmentVariable('Path', 'Machine')
+
+.. code-block:: powershell


Why not put all 3 lines in the same code block?

ThomasXilinx · 2025-10-15T17:05:11Z

docs/inst.rst

+
+.. code-block:: powershell
+
+  $newPaths = "C:\Users\<user>\miniforge3\Scripts;C:\Users\<user>\miniforge3\condabin"


This will only work for miniforge. If people have Anaconda or Miniconda, this will not work.

ThomasXilinx

Some of the proposed changes need more discussion

dwithchenna and others added 8 commits October 8, 2025 16:00

Merge pull request amd#495 from amd/develop

b08189d

merging changes to highlight automatic BF16 conversion

HW support table and LLM Support table

34fa6ef

index and installation updates

8f5bb95

installation readme updates

7c7344a

syntax of bullets

6e34db3

windows install wording

4f8f6d4

added back cmake in table

ddd814c

syntax for title Windows Installation Instructions

26e743d

dwithchenna requested review from ThomasXilinx, Copilot, dwithchenna and uday610 October 15, 2025 03:56

dwithchenna assigned bconsolvo Oct 15, 2025

dwithchenna added the documentation Improvements or additions to documentation label Oct 15, 2025

Copilot AI reviewed Oct 15, 2025

View reviewed changes

ThomasXilinx reviewed Oct 15, 2025

View reviewed changes

docs/index.rst

- 2025

- ☑️

-

* - Ryzen Z2

Copy link

Collaborator

ThomasXilinx Oct 15, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I don't believe this device is supported.

ThomasXilinx reviewed Oct 15, 2025

View reviewed changes

ThomasXilinx requested changes Oct 15, 2025

View reviewed changes

	The Ryzen AI Software packages should now installed in the conda environment created by the installer.
	The Ryzen AI Software packages should now be installed in the conda environment created by the installer.

	- If needed, download the NPU driver version: 32.0.203.280 or the latest 32.0.203.304 here:
	- If needed, download the NPU driver version: 32.0.203.280 or the latest 32.0.203.304 from one of the following links:

		The CIFAR-10 dataset consists of 60,000 32x32 colour images in 10 classes, with 6,000 images per class. There are 50,000 training images and 10,000 test images. You can learn more about the CIFAR-10 dataset here: https://www.cs.toronto.edu/~kriz/cifar.html. This dataset is used in the subsequent steps for quantization and inference. The script also exports the provided PyTorch model into ONNX format. The following snippet from the script shows how the ONNX model is exported:

		The C++ source files, CMake list files, and related artifacts are provided in the ``cpp/resnet_cifar/`` folder. The source file ``cpp/resnet_cifar/resnet_cifar.cpp`` takes 10 images from the CIFAR-10 test set, converts them to .png format, preprocesses them, and performs model inference. The example has onnxruntime dependencies that are provided in ``%RYZEN_AI_INSTALLATION_PATH%/onnxruntime/``.

		Run the following command to build the resnet example. Assign ``-DOpenCV_DIR`` to the OpenCV build directory.



	.. note::
	The following command uses Windows-style backslashes and is intended for use in a Windows environment.


		.. code-block:: powershell

		$newPaths = "C:\Users\<user>\miniforge3\Scripts;C:\Users\<user>\miniforge3\condabin"

overhaul elements of index.rst and inst.rst #497

Are you sure you want to change the base?

overhaul elements of index.rst and inst.rst #497

Uh oh!

Conversation

bconsolvo commented Oct 10, 2025

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull Request Overview

Reviewed Changes

Uh oh!

Copilot AI Oct 15, 2025

Choose a reason for hiding this comment

Uh oh!

Copilot AI Oct 15, 2025

Choose a reason for hiding this comment

Uh oh!

Copilot AI Oct 15, 2025

Choose a reason for hiding this comment

Uh oh!

Copilot AI Oct 15, 2025

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

ThomasXilinx left a comment

Choose a reason for hiding this comment

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants