Arao AI
Back to search results

Qwen3.8-27B · Q4_K_M

Start a discussion
NVIDIA RTX Spark128GBllama.cpp v0.32L2 Cross-framework verified

This page aggregates 1 real-world runs of Qwen3.8-27B (Q4_K_M) on NVIDIA RTX Spark with llama.cpp, contributed by 1 independent source platforms; metrics are averages of published measurements.

10.59 tok/s

Decode

Decode speed

58.05 tok/s

Prefill

Prefill speed

1.19 s

TTFT

Time to first token

48.59 GB

VRAM

VRAM usage

L2 Cross-framework verified

1

Measured runs

1

Independent sources

Other

Source platforms

17 days ago

Last verified

Performance

  1. llama.cpp · Q5_K_M Decode 8.81 · Prefill 37.51 ·
  2. llama.cpp · Q8_0 Decode 6.42 · Prefill 24.36 ·
  3. Ollama · Q5_K_M Decode 7.6 · Prefill 28.29 ·
  4. Ollama · Q6_K Decode 7.47 · Prefill 38.07 ·
  5. Ollama · Q8_0 Decode 6.14 · Prefill 31.39 ·
  6. llama.cpp · Q4_K_M (current)Decode 10.59 · Prefill 58.05 ·
  7. Ollama · Q4_K_M Decode 9.3 · Prefill 41.72 ·
  8. Ollama · Q4_K_S Decode 9.16 · Prefill 49.98 ·

Core figures

Decode (avg)
10.59 tok/s
Prefill (avg)
58.05 tok/s
TTFT (avg)
1.19 s
VRAM (avg)
48.59 GB
MTP acceptance rate
—
TTFB
0.08 GB
Power draw
82.6 W

Configuration

Member-level fields are taken from the most recent run

Model
Qwen3.8-27B
Quantization
Q4_K_M
Framework
llama.cpp
Version
v0.32
Context length
8192 tokens
Batch size
256
GPU layers
99
Flash Attention
Off

Hardware

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

GPU
NVIDIA RTX Spark
Nominal VRAM
128 GB
Measured VRAM (avg)
48.59 GB
OS
Windows 11 24H2
Driver
NVIDIA 590.10
CUDA
13.0
Power draw
82.6 W
Reference price
$2,399 · updated 2026-10-19

Prices are not live; check the store page

Get started

Original model resources

Base model resources may not match this quantization. Use the verified deployment weights above when available.

Sources & evidence

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

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

    v0.32 · Windows 11 24H2 · CUDA 13.0 · 8192 ctx

    10.59 tok/s

    Decode

    58.05 tok/s

    Prefill

    1.19 s

    TTFT

    48.59 GB

    VRAM

    —

    MTP

    82.6 W

    Power draw

    L2 batch seed 2026-10-19: llama.cpp on hw_id=110 running model_id=5 at Q4_K_M