Arao AI
Back to search results

Qwen3.8-9B · IQ1_S

Start a discussion
Intel Arc B58012GBOllama v0.25L2 Cross-framework verified

This page aggregates 1 real-world runs of Qwen3.8-9B (IQ1_S) on Intel Arc B580 with Ollama, contributed by 1 independent source platforms; metrics are averages of published measurements.

60.08 tok/s

Decode

Decode speed

208.14 tok/s

Prefill

Prefill speed

1.77 s

TTFT

Time to first token

10.04 GB

VRAM

VRAM usage

L2 Cross-framework verified

1

Measured runs

1

Independent sources

Other

Source platforms

16 days ago

Last verified

Performance

  1. llama.cpp · IQ2_M Decode 59.78 · Prefill 260.58 ·
  2. Ollama · IQ1_S (current)Decode 60.08 · Prefill 208.14 ·
  3. vLLM · IQ1_S Decode 66.31 · Prefill 217.11 ·
  4. llama.cpp · IQ2_XS Decode 61.39 · Prefill 302.91 ·
  5. llama.cpp · IQ3_XS Decode 56.63 · Prefill 277.66 ·
  6. llama.cpp · Q4_K_M Decode — · Prefill — ·

Core figures

Decode (avg)
60.08 tok/s
Prefill (avg)
208.14 tok/s
TTFT (avg)
1.77 s
VRAM (avg)
10.04 GB
MTP acceptance rate
—
TTFB
0.48 GB
Power draw
158.4 W

Configuration

Member-level fields are taken from the most recent run

Model
Qwen3.8-9B
Quantization
IQ1_S
Framework
Ollama
Version
v0.25
Context length
32768 tokens
Batch size
512
GPU layers
28
Flash Attention
On

Hardware

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

GPU
Intel Arc B580
Nominal VRAM
12 GB
Measured VRAM (avg)
10.04 GB
OS
Windows 11 24H2
Driver
Intel Arc 32.0.101.6552
CUDA
12.6
Power draw
158.4 W

Sources & evidence

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

  1. L2 Cross-framework verifiedOtherOriginal link Verified on 2026-10-19

    v0.25 · Windows 11 24H2 · CUDA 12.6 · 32768 ctx

    60.08 tok/s

    Decode

    208.14 tok/s

    Prefill

    1.77 s

    TTFT

    10.04 GB

    VRAM

    —

    MTP

    158.4 W

    Power draw

    L2 batch seed 2026-10-19: Ollama on hw_id=94 running model_id=52 at IQ1_S