Arao AI
Back to search results

Qwen3-8B · IQ4_XS

Start a discussion
NVIDIA RTX 509032GBStrata v0.37L2 Cross-framework verified

This page aggregates 1 real-world runs of Qwen3-8B (IQ4_XS) on NVIDIA RTX 5090 with Strata, contributed by 1 independent source platforms; metrics are averages of published measurements.

340.69 tok/s

Decode

Decode speed

1,198.68 tok/s

Prefill

Prefill speed

0.14 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

16 days ago

Last verified

Performance

  1. Ollama · IQ3_S Decode 275.39 · Prefill 1404.89 ·
  2. Strata · IQ4_XS (current)Decode 340.69 · Prefill 1198.68 ·
  3. llama.cpp · IQ4_XS Decode 256.42 · Prefill 1150.82 ·
  4. llama.cpp · Q5_K_M Decode 221.44 · Prefill 1381.14 ·
  5. vLLM · IQ3_S Decode 285.95 · Prefill 1301.24 ·
  6. llama.cpp · Q6_K Decode 184.6 · Prefill 725.28 ·
  7. vLLM · IQ2_M Decode 325.14 · Prefill 1583.07 ·
  8. vLLM · Q2_K Decode 357.71 · Prefill 1294.78 ·
  9. vLLM · IQ4_XS Decode 252.51 · Prefill 829.1 ·
  10. Ollama · Q5_K_M Decode 172.9 · Prefill 940.17 ·
  11. llama.cpp · Q8_0 Decode 162.5 · Prefill 1218.8 ·
  12. llama.cpp · Q4_K_M Decode 285.3 · Prefill 2139.8 ·

Core figures

Decode (avg)
340.69 tok/s
Prefill (avg)
1,198.68 tok/s
TTFT (avg)
0.14 s
VRAM (avg)
14.53 GB
MTP acceptance rate
0.77
TTFB
2.73 GB
Power draw
412.2 W

Configuration

Member-level fields are taken from the most recent run

Model
Qwen3-8B
Quantization
IQ4_XS
Framework
Strata
Version
v0.37
Context length
8192 tokens
Batch size
512
GPU layers
15
Flash Attention
On

Hardware

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

GPU
NVIDIA RTX 5090
Nominal VRAM
32 GB
Measured VRAM (avg)
14.53 GB
OS
Windows 11 24H2
Driver
570.86
CUDA
12.4
Power draw
412.2 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

    v0.37 · Windows 11 24H2 · CUDA 12.4 · 8192 ctx

    340.69 tok/s

    Decode

    1,198.68 tok/s

    Prefill

    0.14 s

    TTFT

    14.53 GB

    VRAM

    0.77

    MTP

    412.2 W

    Power draw

    L2 batch seed 2026-10-19: Strata on hw_id=13 running model_id=53 at IQ4_XS