Arao AI
Back to search results

Qwen3-8B · IQ4_XS

Start a discussion
NVIDIA RTX 309024GBOllama v0.54L2 Cross-framework verified

This page aggregates 1 real-world runs of Qwen3-8B (IQ4_XS) on NVIDIA RTX 3090 with Ollama, contributed by 1 independent source platforms; metrics are averages of published measurements.

114.83 tok/s

Decode

Decode speed

486.18 tok/s

Prefill

Prefill speed

0.67 s

TTFT

Time to first token

14.53 GB

VRAM

VRAM usage

L2 Cross-framework verified

1

Measured runs

1

Independent sources

Other

Source platforms

Today

Last verified

Performance

  1. Ollama · IQ4_XS (current)Decode 114.83 · Prefill 486.18 ·
  2. Strata · Q5_K_S Decode 138.04 · Prefill 495.5 ·
  3. llama.cpp · Q8_0 Decode 84.9 · Prefill 636.8 ·
  4. llama.cpp · Q5_K_M Decode 123.7 · Prefill 927.8 ·
  5. llama.cpp · Q4_K_M Decode 149 · Prefill 1117.5 ·

Core figures

Decode (avg)
114.83 tok/s
Prefill (avg)
486.18 tok/s
TTFT (avg)
0.67 s
VRAM (avg)
14.53 GB
MTP acceptance rate
—
TTFB
0.92 GB
Power draw
313.6 W

Configuration

Member-level fields are taken from the most recent run

Model
Qwen3-8B
Quantization
IQ4_XS
Framework
Ollama
Version
v0.54
Context length
16384 tokens
Batch size
256
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
NVIDIA RTX 3090
Nominal VRAM
24 GB
Measured VRAM (avg)
14.53 GB
OS
macOS 15.0
Driver
560.94
CUDA
12.6
Power draw
313.6 W

Sources & evidence

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

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

    v0.54 · macOS 15.0 · CUDA 12.6 · 16384 ctx

    114.83 tok/s

    Decode

    486.18 tok/s

    Prefill

    0.67 s

    TTFT

    14.53 GB

    VRAM

    —

    MTP

    313.6 W

    Power draw

    L2 batch seed 2026-10-10: Ollama on hw_id=2 running model_id=53 at IQ4_XS