Getting Started

This page gets you from a bare Linux install to a working ROCm + PyTorch stack on a gfx1030 card, and explains how to preview this wiki locally.

1. Confirm your hardware

Make sure your card is a Navi 21 gfx1030 (or a related RDNA2 card):

lspci | grep -i vga
# After ROCm is installed:
rocminfo | grep -m1 -o 'gfx[0-9]*'

See Supported Hardware for the full card list.

2. Install ROCm

Follow Installing ROCm. At a high level:

# Add the amdgpu repo, then:
sudo apt install rocm
sudo usermod -aG render,video "$LOGNAME"
# Reboot, then verify:
rocminfo
clinfo | grep -i 'gfx\|Board'

If you run one or more Radeon PRO V620 (or PRO W6800) cards, these tweaks are worth doing before you load models (Fedora or Ubuntu 26.04 for power cap; P2P is Fedora + AMD-validated — A/B on Intel, see When P2P helps):

Power Tuning — drop the VBIOS-locked 250 W floor to 120 W and boot-cap at 180 W. Passthrough VM and need OD clocks instead? See Soft unlock (amd-v620-soft-unlock).
Disabling ECC — Pro cards hide ~2 GB behind ECC; optional extra VRAM.
Multi-GPU PCIe P2P — enable GPU↔GPU peer-to-peer for multi-card setups.

Baremetal power floor + P2P come from the v620_toolbox repo.

4. Run an inference stack (Docker)

The fastest path is the prebuilt images — no local ROCm/PyTorch/vLLM build required:

Running vLLM (Docker) — blivioniag/vllm-rdna on a blivioniag/rocm-rdna PyTorch base.
vLLM recipes — Hub -extras vs recipe container vs Flash-Next by card count.

Prefer GGUF and building from source? See Building & Running llama.cpp (ROCm or Vulkan).

Want to build the images yourself, or use the RDNA-tuned kernels? See Building the Images and The rdna_extras fork.

5. Smoke test

# Inside a rocm-rdna / vllm-rdna container, or a local ROCm PyTorch env:
python -c "import torch; print(torch.cuda.is_available(), torch.cuda.get_device_name(0))"

You should see True and your Radeon card’s name. If not, see Troubleshooting.

6. Join the community Discord

Questions, live benches, and fork updates land first on gfx1030 club: https://discord.gg/mESex2aBp (#vllm-rdna, #llamacpp, #general, #benchmarks). See Useful resources.

Previewing this wiki locally

This site is built with mdBook. You do not need a GPU or ROCm to work on the docs.

Prerequisites

Git — to clone the repository.
mdBook — a single static binary (no runtime dependencies).

Install mdBook

The quickest way is to grab a prebuilt binary from the mdBook releases page:

mkdir -p "$HOME/.local/bin"
MDBOOK_VERSION=v0.5.4
curl -sL "https://github.com/rust-lang/mdBook/releases/download/${MDBOOK_VERSION}/mdbook-${MDBOOK_VERSION}-x86_64-unknown-linux-gnu.tar.gz" \
  | tar -xz -C "$HOME/.local/bin"
export PATH="$HOME/.local/bin:$PATH"
mdbook --version

If you have a Rust toolchain, cargo install mdbook also works.

Build and preview

git clone https://github.com/blivioniag/gfx1030-wiki.git
cd gfx1030-wiki

mdbook build      # outputs static HTML to ./book
mdbook serve      # live-reloading preview at http://localhost:3000

Add content

Create a new markdown file in src/, e.g. src/my_page.md.
Add an entry for it in src/SUMMARY.md.
Re-run mdbook serve to preview.
Open a pull request against master.

See Contributing for the full guidelines.

Wiki GFX1030