AMD Strix Halo (gfx1151) — vLLM Toolbox/Container

An Fedora 43 Docker/Podman container that is Toolbx-compatible (usable as a Fedora toolbox) for serving LLMs with vLLM on AMD Ryzen AI Max “Strix Halo” (gfx1151). Built on the TheRock nightly builds for ROCm.

---

🚀 High-Performance Clustering Support (New!)

Update: This toolbox now ships with a custom build of ROCm/RCCL that enables native RDMA/RoCE v2 support for Strix Halo (gfx1151). This allows you to connect two nodes via a low-latency interconnect (e.g., Intel E810) and run vLLM with Tensor Parallelism (TP=2) effectively acting as a single 256GB Unified Memory GPU.

👉 Read the Full RDMA Cluster Setup Guide for hardware requirements and configuration instructions.

---

📦 Project Context

This repository is part of the Strix Halo AI Toolboxes project. Check out the website for an overview of all toolboxes, tutorials, and host configuration guides.

❤️ Support

This is a hobby project maintained in my spare time. If you find these toolboxes and tutorials useful, you can buy me a coffee to support the work! ☕

🙏 Acknowledgments

• Adrian (@Lafunamor): Huge thanks for all the help, PRs, and testing to get this project stabilized!
• Patrick Audley (paudley/ai-notes): Thanks for the strix-halo build notes. This toolbox relies on that research (specifically the Triton patches and aiter compilation strategy) to successfully run vLLM and AITER Flash-Attention on Strix Halo.

---

Table of Contents

• Tested Models (Benchmarks)
• 1) Toolbx vs Docker/Podman
• 2) Quickstart — Fedora Toolbx
• 3) Quickstart — Ubuntu (Distrobox)
• 4) Testing the API
• 5) Use a Web UI for Chatting
• 6) Host Configuration
• 7) Distributed Clustering (RDMA/RoCE)

Tested Models (Benchmarks)

Important

Note on Throughput: These benchmarks measure Peak Multi-User Throughput (Tokens/Second) at high concurrency (batching multiple sequences simultaneously to saturate the Strix Halo's memory bandwidth). If you are testing with a single request (Concurrency = 1), your individual generation speed will be lower than these maximum hardware-saturation numbers. These metrics represent the total capacity of the system under heavy load.

View full benchmarks at: https://kyuz0.github.io/amd-strix-halo-vllm-toolboxes/

Model	Params / Quant	GPU Requirement
meta-llama/Meta-Llama-3.1-8B-Instruct	8B / BF16	1 GPU (TP=1, 2)
google/gemma-4-26B-A4B-it	26B / BF16	1 GPU (TP=1, 2)
google/gemma-4-31B-it	31B / BF16	1 GPU (TP=1, 2)
openai/gpt-oss-20b	20B / BF16	1 GPU (TP=1, 2)
openai/gpt-oss-120b	120B / BF16	1 GPU (TP=1)
Qwen/Qwen3.6-35B-A3B	35B / BF16	1 GPU (TP=1)
cyankiwi/Qwen3.6-35B-A3B-AWQ-4bit	35B / AWQ 4-bit	1 GPU (TP=1)
cyankiwi/Qwen3.5-122B-A10B-AWQ-4bit	122B / AWQ 4-bit	1 GPU (TP=1, 2)
cyankiwi/Qwen3.5-122B-A10B-AWQ-8bit	122B / AWQ 8-bit	2 GPUs (TP=2 Only)
cyankiwi/MiniMax-M2.7-AWQ-4bit	N/A / AWQ 4-bit	2 GPUs (TP=2 Only)
ayysasha/MiniMax-M2.7-AWQ-G32-STRIX-2H	N/A / Mixed BF16+INT4 AWQ	2 GPUs (TP=2 Only)

---

1) Toolbx vs Docker/Podman

The kyuz0/vllm-therock-gfx1151 image is available in two channels:

Tag	Description
:latest	Last verified working build. Recommended for most users.
:dev	Absolute latest build. May contain upstream regressions.

The image can be used both as:

• Fedora Toolbx (recommended for development): Toolbx shares your HOME and user, so models/configs live on the host. Great for iterating quickly while keeping the host clean.
• Docker/Podman (recommended for deployment/perf): Use for running vLLM as a service (host networking, IPC tuning, etc.). Always mount a host directory for model weights so they stay outside the container.

---

2) Quickstart — Fedora Toolbx

Recommended: Use the included refresh_toolbox.sh script. It pulls the image and creates the toolbox with the correct parameters:

# Interactive — prompts you to choose latest (default) or dev
./refresh_toolbox.sh

# Or specify directly:
./refresh_toolbox.sh latest   # verified working build
./refresh_toolbox.sh dev      # bleeding edge

InfiniBand / RDMA Support: The script automatically detects if a fast InfiniBand link is active (checks /dev/infiniband). If found, it correctly sets up the container to expose these devices, enabling high-performance clustering.

Manual Creation:

To manually create a toolbox that exposes the GPU and relaxes seccomp:

toolbox create vllm \
  --image docker.io/kyuz0/vllm-therock-gfx1151:latest \
  -- --device /dev/dri --device /dev/kfd \
  --group-add video --group-add render --security-opt seccomp=unconfined

Enter it:

toolbox enter vllm

Model storage: Models are downloaded to ~/.cache/huggingface by default. This directory is shared with the host if you created the toolbox correctly, so downloads persist.

Serving a Model (Easiest Way)

The toolbox includes a TUI wizard called start-vllm which includes pre-configured models and handles the launch flags for you. This is the easiest way to get started.

start-vllm

Cache note: vLLM writes compiled kernels to ~/.cache/vllm/.

---

3) Quickstart — Ubuntu (Distrobox)

Ubuntu’s toolbox package still breaks GPU access, so use Distrobox instead:

distrobox create -n vllm \
  --image docker.io/kyuz0/vllm-therock-gfx1151:latest \
  --additional-flags "--device /dev/kfd --device /dev/dri --group-add video --group-add render --security-opt seccomp=unconfined"

distrobox enter vllm

Verification: Run rocm-smi to check GPU status.

Serving a Model (Easiest Way)

The toolbox includes a TUI wizard called start-vllm which includes pre-configured models and handles the launch flags for you. This is the easiest way to get started.

start-vllm

---

4) Testing the API

Once the server is up, hit the OpenAI‑compatible endpoint:

curl -X POST http://localhost:8000/v1/chat/completions \
  -H "Content-Type: application/json" \
  -d '{"model":"Qwen/Qwen2.5-7B-Instruct","messages":[{"role":"user","content":"Hello! Test the performance."}]}'

You should receive a JSON response with a choices[0].message.content reply.

If you don't want to bother specifying the model name, you can run this which will query the currently deployed model:

MODEL=$(curl -s http://localhost:8000/v1/models | jq -r '.data[0].id') curl -X POST http://localhost:8000/v1/chat/completions \
  -H "Content-Type: application/json" \
  -d "{
    \"model\": \"$MODEL\",
    \"messages\":[{\"role\":\"user\",\"content\":\"Hello! Test the performance.\"}]
  }"

---

5) Use a Web UI for Chatting

If vLLM is on a remote server, expose port 8000 via SSH port forwarding:

ssh -L 0.0.0.0:8000:localhost:8000 <vllm-host>

Then, you can start HuggingFace ChatUI like this (on your host):

docker run -p 3000:3000 \
  --add-host=host.docker.internal:host-gateway \
  -e OPENAI_BASE_URL=http://host.docker.internal:8000/v1 \
  -e OPENAI_API_KEY=dummy \
  -v chat-ui-data:/data \
  ghcr.io/huggingface/chat-ui-db

6) Host Configuration

This should work on any Strix Halo. For a complete list of available hardware, see: Strix Halo Hardware Database

6.1 Test Configuration

Component	Specification
Test Machine	Framework Desktop
CPU	Ryzen AI MAX+ 395 "Strix Halo"
System Memory	128 GB RAM
GPU Memory	512 MB allocated in BIOS
Host OS	Fedora 43, Linux 6.18.5-200.fc43.x86_64

6.2 Kernel Parameters (tested on Fedora 42)

Add these boot parameters to enable unified memory while reserving a minimum of 4 GiB for the OS (max 124 GiB for iGPU):

Warning

Based on benchmarking by Lars Urban (@urbanswelt), there is definitive indication that setting amd_iommu=off performs better than the previously recommended iommu=pt. Key result: amd_iommu=off is 5-12% faster than either IOMMU-enabled mode. See Issue #66 for details.

amd_iommu=off amdgpu.gttsize=126976 ttm.pages_limit=32505856

Parameter	Purpose
amd_iommu=off	Disables the AMD IOMMU. This improves performance over iommu=pt, reducing overhead for both the RDMA NIC and the iGPU unified memory access.
amdgpu.gttsize=126976	Caps GPU unified memory to 124 GiB; 126976 MiB ÷ 1024 = 124 GiB
ttm.pages_limit=32505856	Caps pinned memory to 124 GiB; 32505856 × 4 KiB = 126976 MiB = 124 GiB

Source: https://www.reddit.com/r/LocalLLaMA/comments/1m9wcdc/comment/n5gf53d/?context=3&utm_source=share&utm_medium=web3x&utm_name=web3xcss&utm_term=1&utm_content=share_button

Apply the changes:

# Edit /etc/default/grub to add parameters to GRUB_CMDLINE_LINUX
sudo grub2-mkconfig -o /boot/grub2/grub.cfg
sudo reboot

7) Distributed Clustering (RDMA/RoCE)

This toolbox supports high-performance clustering of multiple Strix Halo nodes using Infiniband or RoCE v2 (e.g., Intel E810). This enables Tensor Parallelism across machines with extremely low latency (~5µs).

Detailed Documentation: RDMA Cluster Setup Guide

Key Features:

• Custom RCCL Patch: Use of a custom-built librccl.so to support RDMA on gfx1151.
• Easy Setup: refresh_toolbox.sh automatically detects and exposes RDMA devices.
• Cluster Management: Included start-vllm-cluster TUI for managing Ray and vLLM.

8) AITER on Strix Halo Support Status

This toolbox supports running AITER Flash Attention on Strix Halo (gfx1151). Normally, vLLM crashes on RDNA APUs if VLLM_ROCM_USE_AITER=1 is enabled, because AITER attempts to JIT-compile CDNA-specific MoE (Mixture of Experts) and CustomOps assembly instructions that lack RDNA hardware support.

To bypass this limitation, scripts/patch_strix.py applies a few APU-specific guards (building on the work from ai-notes linked above):

• Patch 2 (vllm/_aiter_ops.py): Intercepts the MoE gate (is_fused_moe_enabled()) forcing it to disable AITER MoE and Linear FP8 on gfx1x architectures.
• Patch 3.5 (vllm/model_executor/layers/fused_moe/oracle/unquantized.py): Blocks the VLLM_ROCM_USE_AITER_MOE environment variable from forcing a JIT compile override.
• Patch 5 (vllm/platforms/rocm.py): Bypasses the RMSNorm custom op registration on gfx1x to prevent CUDA Graph capture crashes during model initialization.

Because of these patches, when ROCm Attention is selected in the launcher, vLLM routes Attention to AITER (using the ds_swizzle RDNA header fallbacks injected via scripts/patch_aiter_headers.py), while safely falling back to Triton for MoE matrices and Torch/Triton for RMSNorm.