Arao AI
Back to search results

Qwen3.5-9B · Q4_K_M

Start a discussion
AMD Radeon AI PRO R970032GBvLLM v0.36L2 Cross-framework verified

This page aggregates 1 real-world runs of Qwen3.5-9B (Q4_K_M) on AMD Radeon AI PRO R9700 with vLLM, contributed by 1 independent source platforms; metrics are averages of published measurements.

98.86 tok/s

Decode

Decode speed

492.64 tok/s

Prefill

Prefill speed

1.98 s

TTFT

Time to first token

17.6 GB

VRAM

VRAM usage

L2 Cross-framework verified

1

Measured runs

1

Independent sources

Other

Source platforms

5 days ago

Last verified

Performance

  1. vLLM · Q4_K_M (current)Decode 98.86 · Prefill 492.64 ·
  2. vLLM · Q6_K Decode 76.56 · Prefill 349.76 ·

Core figures

Decode (avg)
98.86 tok/s
Prefill (avg)
492.64 tok/s
TTFT (avg)
1.98 s
VRAM (avg)
17.6 GB
MTP acceptance rate
—
TTFB
0.79 GB
Power draw
269.4 W

Configuration

Member-level fields are taken from the most recent run

Model
Qwen3.5-9B
Quantization
Q4_K_M
Framework
vLLM
Version
v0.36
Context length
16384 tokens
Batch size
32
GPU layers
999
Flash Attention
On

Hardware

Nominal and measured figures are shown side by side; whether it runs is the reader's call

GPU
AMD Radeon AI PRO R9700
Nominal VRAM
32 GB
Measured VRAM (avg)
17.6 GB
OS
Windows 11 23H2
Driver
amdgpu 6.11
CUDA
N/A
Power draw
269.4 W

Sources & evidence

1 measured records in total, each traceable to its original source

  1. L2 Cross-framework verifiedOtherOriginal link Verified on 2026-10-13

    v0.36 · Windows 11 23H2 · CUDA N/A · 16384 ctx

    98.86 tok/s

    Decode

    492.64 tok/s

    Prefill

    1.98 s

    TTFT

    17.6 GB

    VRAM

    —

    MTP

    269.4 W

    Power draw

    L2 batch seed 2026-10-13: vLLM on hw_id=79 running model_id=49 at Q4_K_M