vLLM/recipes/models/Qwen/Qwen3-30B-A3B.yaml

meta:
  title: "Qwen3-30B-A3B"
  slug: "qwen3-30b-a3b"
  provider: "Qwen"
  description: "Qwen3 MoE model with 30.5B total and 3.3B active parameters, validated for BF16 serving on Intel Xeon 6 and Intel Arc Pro B70 and W8A8 serving on Ascend A3."
  date_added: 2026-08-14
  date_updated: 2026-09-23
  tasks:
    - text
  performance_headline: "Validated on Intel Xeon 6 (GNR), Ascend A3 W8A8 and 4x Intel Arc Pro B70 with BF16 serving"
  related_recipes:
    - "Qwen/Qwen3-235B-A22B-Instruct-2507"
  hardware:
    xeon6: verified
    ascend_910c: verified
    arc_pro_b70: verified

model:
  model_id: "Qwen/Qwen3-30B-A3B"
  min_vllm_version: "0.8.5"
  architecture: moe
  parameter_count: "30.5B"
  active_parameters: "3.3B"
  context_length: 40960
  base_args: []
  base_env: {}

features:
  tool_calling:
    description: "Enable automatic tool choice with the Hermes parser"
    args:
      - "--enable-auto-tool-choice"
      - "--tool-call-parser"
      - "hermes"
  reasoning:
    description: "Enable Qwen3 thinking mode with the Qwen3 reasoning parser"
    args:
      - "--reasoning-parser"
      - "qwen3"
  long_context:
    description: "YaRN RoPE scaling to the model card's validated 131,072-token context. Static YaRN holds the scaling factor constant regardless of input length, so leave this off unless you actually serve long prompts — it can cost quality below 32K."
    args:
      - "--hf-overrides"
      - '{"max_position_embeddings":131072,"rope_parameters":{"rope_type":"yarn","rope_theta":1000000,"factor":4.0,"original_max_position_embeddings":32768}}'
      - "--max-model-len"
      - "131072"

opt_in_features:
  - long_context

hardware_opt_in_features:
  ascend_910c:
    - tool_calling
    - reasoning

variants:
  default:
    precision: bf16
    vram_minimum_gb: 74
    description: "Full precision BF16"
    supported_hardware: [xeon6, arc_pro_b70]
  ascend_w8a8:
    model_id: "Eco-Tech/Qwen3-30B-A3B-w8a8"
    precision: int8
    label: "W8A8 (Ascend A3)"
    vram_minimum_gb: 37
    description: "ModelSlim INT8 W8A8 checkpoint for Ascend 910C, verified with TP4 + expert parallelism."
    supported_hardware: [ascend_910c]
    tp: 4
    min_vllm_version: "0.23.0"
    extra_args:
      - "--trust-remote-code"
      - "--max-num-seqs"
      - "100"
      - "--max-model-len"
      - "40960"
      - "--max-num-batched-tokens"
      - "16384"
      - "--enable-expert-parallel"
      - "--quantization"
      - "ascend"
      - "--distributed_executor_backend"
      - "mp"
      - "--no-enable-prefix-caching"
      - "--async-scheduling"
      - "--compilation-config"
      - '{"cudagraph_mode": "FULL_DECODE_ONLY"}'
      - "--additional-config"
      - '{"enable_flashcomm1": true, "weight_nz_mode": 2}'
      - "--gpu-memory-utilization"
      - "0.95"
    extra_env:
      VLLM_USE_MODELSCOPE: "True"
      HCCL_OP_EXPANSION_MODE: "AIV"
      HCCL_BUFFSIZE: "1024"
      OMP_PROC_BIND: "false"
      OMP_NUM_THREADS: "1"
      TASK_QUEUE_ENABLE: "1"
      PYTORCH_NPU_ALLOC_CONF: "expandable_segments:True"
      VLLM_USE_RUST_FRONTEND: "1"
    hardware_overrides:
      ascend_910c:
        docker_image: "swr.cn-southwest-2.myhuaweicloud.com/base_image/ascend-ci/vllm-ascend:v0.23.0-a3"

compatible_strategies:
  - single_node_tp
  - multi_node_tp

hardware_overrides:
  cpu:
    extra_args:
      - "--max-num-batched-tokens"
      - "16384"
      - "--gpu-memory-utilization"
      - "0.8"
  xpu:
    extra_args:
      - "--block-size"
      - "64"
      - "--enforce-eager"
      - "--no-enable-prefix-caching"
      - "--disable-sliding-window"
      - "--max-model-len"
      - "9472"
      - "--max-num-batched-tokens"
      - "8192"
      - "--gpu-memory-utilization"
      - "0.9"

strategy_overrides: {}

guide: |
  ## Overview

  Qwen3-30B-A3B is a mixture-of-experts Qwen3 model with 30.5B total parameters
  and 3.3B active parameters per token. This recipe includes an Intel Xeon 6 CPU
  configuration validated with BF16 serving and an Ascend A3 W8A8 variant.

  ## Prerequisites

  - Hardware: 4x Xeon 6 CPUs, or 4x Intel Arc Pro B70 (32 GB per card).
  - vLLM >= 0.8.5

  ### pip (Intel Xeon 6 CPUs)

  For Intel and AMD x86 CPUs, follow the
  [CPU pre-built wheels](https://docs.vllm.ai/en/latest/getting_started/installation/cpu/#pre-built-wheels)
  installation instructions.

  ### Docker (Intel Xeon 6 CPUs)

  ```bash
  docker pull vllm/vllm-openai-cpu:latest-x86_64
  ```

  ### Docker (Intel Arc Pro B70)

  ```bash
  docker pull vllm/vllm-openai-xpu:latest
  ```

  ## Intel Xeon 6

  Choose the tensor parallel size based on the system topology and deployment
  requirements. Add `--tensor-parallel-size <N>` when needed; the recipe does
  not prescribe a Xeon 6 TP value or hard-code NUMA node IDs.

  ```bash
  vllm serve Qwen/Qwen3-30B-A3B \
    --max-num-batched-tokens 16384 \
    --gpu-memory-utilization 0.8
  ```

  Docker:

  ```bash
  docker run \
    --privileged --ipc=host -p 8000:8000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    vllm/vllm-openai-cpu:latest-x86_64 Qwen/Qwen3-30B-A3B \
    --tensor-parallel-size 1 \
    --max-num-batched-tokens 16384 \
    --gpu-memory-utilization 0.8 \
    --enable-auto-tool-choice \
    --tool-call-parser hermes \
    --reasoning-parser qwen3
  ```

  ## Intel Arc Pro B70 (XPU)

  Validated on 4x Intel Arc Pro B70 (32 GB per card) with the official vLLM XPU
  image `vllm/vllm-openai-xpu:latest`, BF16, TP=4.

  ```bash
  docker run --device /dev/dri \
    -v /dev/dri/by-path:/dev/dri/by-path --shm-size=16g \
    --privileged --ipc=host -p 8000:8000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    vllm/vllm-openai-xpu:latest \
    Qwen/Qwen3-30B-A3B \
    --tensor-parallel-size 4 \
    --block-size 64 \
    --enforce-eager \
    --no-enable-prefix-caching \
    --disable-sliding-window \
    --max-model-len 9472 \
    --max-num-batched-tokens 8192 \
    --gpu-memory-utilization 0.9
  ```

  ## Client Usage

  ```python
  from openai import OpenAI

  client = OpenAI(base_url="http://localhost:8000/v1", api_key="unused")
  response = client.chat.completions.create(
      model="Qwen/Qwen3-30B-A3B",
      messages=[{"role": "user", "content": "Explain tensor parallelism briefly."}],
  )
  print(response.choices[0].message.content)
  ```

  ## References

  - [Model card](https://huggingface.co/Qwen/Qwen3-30B-A3B)
  - [Ascend A3 W8A8 checkpoint](https://www.modelscope.cn/models/Eco-Tech/Qwen3-30B-A3B-w8a8)
  - [vLLM Ascend](https://github.com/vllm-project/vllm-ascend)
  - [vLLM CPU installation](https://docs.vllm.ai/en/latest/getting_started/installation/cpu/)