Arao AI
Back to search results

Qwen3.5-9B · IQ3_XS

Start a discussion
AMD Strix Point (Radeon 890M)Ollama v0.45L2 Cross-framework verified

This page aggregates 1 real-world runs of Qwen3.5-9B (IQ3_XS) on AMD Strix Point (Radeon 890M) with Ollama, contributed by 1 independent source platforms; metrics are averages of published measurements.

62.85 tok/s

Decode

Decode speed

248.56 tok/s

Prefill

Prefill speed

1.31 s

TTFT

Time to first token

13.82 GB

VRAM

VRAM usage

L2 Cross-framework verified

1

Measured runs

1

Independent sources

Other

Source platforms

2 days ago

Last verified

Performance

  1. vLLM · IQ3_S Decode 65.94 · Prefill 339.59 ·
  2. Ollama · IQ3_XS (current)Decode 62.85 · Prefill 248.56 ·
  3. Ollama · IQ4_XS Decode 46.77 · Prefill 176.41 ·

Core figures

Decode (avg)
62.85 tok/s
Prefill (avg)
248.56 tok/s
TTFT (avg)
1.31 s
VRAM (avg)
13.82 GB
MTP acceptance rate
—
TTFB
0.5 GB
Power draw
56.3 W

Configuration

Member-level fields are taken from the most recent run

Model
Qwen3.5-9B
Quantization
IQ3_XS
Framework
Ollama
Version
v0.45
Context length
2048 tokens
Batch size
512
GPU layers
33
Flash Attention
On

Hardware

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

GPU
AMD Strix Point (Radeon 890M)
Measured VRAM (avg)
13.82 GB
OS
Ubuntu 24.04
Driver
amdgpu-pro 24.20
CUDA
N/A
Power draw
56.3 W

Sources & evidence

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

  1. L2 Cross-framework verifiedOtherOriginal link Verified on 2026-10-16

    v0.45 · Ubuntu 24.04 · CUDA N/A · 2048 ctx

    62.85 tok/s

    Decode

    248.56 tok/s

    Prefill

    1.31 s

    TTFT

    13.82 GB

    VRAM

    —

    MTP

    56.3 W

    Power draw

    L2 batch seed 2026-10-16: Ollama on hw_id=89 running model_id=49 at IQ3_XS