Skip to content

Add documentation missing information for AMD devices. #8546

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
wants to merge 3 commits into
base: main
Choose a base branch
from
Open

Add documentation missing information for AMD devices. #8546

Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
1912bdd
Add documentation missing information for AMD devices.
ArKam Sep 8, 2025
851d44e
Update docs/installation/docker.md
ArKam Sep 13, 2025
ffe5e9c
Merge branch 'main' into fix_rocm_documentation
ArKam Sep 13, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
29 changes: 29 additions & 0 deletions docs/installation/docker.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -22,6 +22,35 @@ Ensure your Docker setup is able to use your GPU. Then:

Once the container starts up, open <http://localhost:9090> in your browser, install some models, and start generating.

## Docker image for AMD ROCm devices

!!! warning "MISSING RENDER_GROUP_ID VARIABLE"

Because of the way the invoke wrapper command script do work, it can't automatically detect your host devices render group.
In order for invoke wrapper command to successfully detect your AMD devices, you need to give it the host render group_id:
```bash
docker run -e "RENDER_GROUP_ID=$(getent group render | cut -d':' -f 3)
```

AMD GPUs users need to get **BOTH** `kfd` and `dri` devices mounted on their container, otherwise Invoke won't be able to detect them properly:

```bash
docker run -e "RENDER_GROUP_ID=<YOUR_HOST_RENDER_GROUP_ID>" --device /dev/kfd --device /dev/dri ghcr.io/invoke-ai/invokeai:v6.4.0-rocm
```
!!! tip "KFD/DRI DEVICES"

`kfd` device created by the AMD linux kernel module, which provides your container with the ability for an application to use specific ROCm/HSA components such as the HSA Runtime and send them computation kernels.
`dri` device created by the AMD linux kernel module, which provides your container the ability to control the GPU card rendering interface.
For more information about HSA Kernels and HSA architectures, look at AMD [documentation](https://github.com/HSAFoundation/HSA-Runtime-AMD)
For more information about DRI (Direct Rendering Interface) framework, look at FreeDesktop/MESA/Khronos [documentation](https://dri.freedesktop.org/wiki/)

!!! tip "AMD LEGACY RX CARDS"

AMD GPUs such as RX6600 can be used by passing passing additional specific `HCC_AMDGPU_TARGET=gfx1030` and `HSA_OVERRIDE_GFX_VERSION=10.3.0` environment variables:
```bash
docker run -e "HCC_AMDGPU_TARGET=gfx1030" -e "HSA_OVERRIDE_GFX_VERSION=10.3.0" -e "RENDER_GROUP_ID=<YOUR_HOST_RENDER_GROUP_ID>" --device /dev/kfd --device /dev/dri ghcr.io/invoke-ai/invokeai:v6.4.0-rocm
```

## Build-It-Yourself

All the docker materials are located inside the [docker](https://github.com/invoke-ai/InvokeAI/tree/main/docker) directory in the Git repo.
Expand Down