Arao AI
Back to search results

Qwen3.5-0.8B · IQ2_XS

Start a discussion
Intel Arc A770 16GB16GBvLLM v0.27L2 Cross-framework verified

This page aggregates 1 real-world runs of Qwen3.5-0.8B (IQ2_XS) on Intel Arc A770 16GB with vLLM, contributed by 1 independent source platforms; metrics are averages of published measurements.

182.37 tok/s

Decode

Decode speed

880.77 tok/s

Prefill

Prefill speed

1.38 s

TTFT

Time to first token

2.91 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 · IQ2_XS (current)Decode 182.37 · Prefill 880.77 ·
  2. vLLM · Q8_0 Decode 87.79 · Prefill 281.11 ·
  3. Strata · IQ2_M Decode 232.67 · Prefill 822.78 ·
  4. Strata · IQ3_XS Decode 236.84 · Prefill 822.88 ·
  5. vLLM · Q3_K_L Decode 146.32 · Prefill 525.63 ·
  6. vLLM · Q5_K_M Decode 109.82 · Prefill 577.25 ·
  7. Ollama · Q2_K Decode 163.98 · Prefill 697.28 ·

Core figures

Decode (avg)
182.37 tok/s
Prefill (avg)
880.77 tok/s
TTFT (avg)
1.38 s
VRAM (avg)
2.91 GB
MTP acceptance rate
—
TTFB
1.46 GB
Power draw
194.8 W

Configuration

Member-level fields are taken from the most recent run

Model
Qwen3.5-0.8B
Quantization
IQ2_XS
Framework
vLLM
Version
v0.27
Context length
16384 tokens
Batch size
8
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
Intel Arc A770 16GB
Nominal VRAM
16 GB
Measured VRAM (avg)
2.91 GB
OS
Windows 11 23H2
Driver
Intel Arc 31.0.101.5186
CUDA
12.6
Power draw
194.8 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

    v0.27 · Windows 11 23H2 · CUDA 12.6 · 16384 ctx

    182.37 tok/s

    Decode

    880.77 tok/s

    Prefill

    1.38 s

    TTFT

    2.91 GB

    VRAM

    —

    MTP

    194.8 W

    Power draw

    L2 batch seed 2026-10-16: vLLM on hw_id=97 running model_id=51 at IQ2_XS