Arao AI
Back to search results

Qwen3-1.7B · IQ4_XS

Start a discussion
Qualcomm Snapdragon X ElitevLLM v0.29L2 Cross-framework verified

This page aggregates 1 real-world runs of Qwen3-1.7B (IQ4_XS) on Qualcomm Snapdragon X Elite with vLLM, contributed by 1 independent source platforms; metrics are averages of published measurements.

100.93 tok/s

Decode

Decode speed

494.53 tok/s

Prefill

Prefill speed

1.61 s

TTFT

Time to first token

4.75 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 · IQ4_XS (current)Decode 100.93 · Prefill 494.53 ·
  2. Ollama · Q3_K_M Decode 88.26 · Prefill 401.25 ·

Core figures

Decode (avg)
100.93 tok/s
Prefill (avg)
494.53 tok/s
TTFT (avg)
1.61 s
VRAM (avg)
4.75 GB
MTP acceptance rate
—
TTFB
0.81 GB
Power draw
34.4 W

Configuration

Member-level fields are taken from the most recent run

Model
Qwen3-1.7B
Quantization
IQ4_XS
Framework
vLLM
Version
v0.29
Context length
32768 tokens
Batch size
32
GPU layers
999
Flash Attention
Off

Hardware

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

GPU
Qualcomm Snapdragon X Elite
Measured VRAM (avg)
4.75 GB
OS
Windows 11 23H2
Driver
Qualcomm Adreno 7.0
CUDA
N/A
Power draw
34.4 W

Sources & evidence

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

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

    v0.29 · Windows 11 23H2 · CUDA N/A · 32768 ctx

    100.93 tok/s

    Decode

    494.53 tok/s

    Prefill

    1.61 s

    TTFT

    4.75 GB

    VRAM

    —

    MTP

    34.4 W

    Power draw

    L2 batch seed 2026-10-16: vLLM on hw_id=104 running model_id=55 at IQ4_XS