Arao AI
Back to search results

Qwen3.8-9B · IQ2_M

Start a discussion
NVIDIA RTX 4080 Super16GBvLLM v0.32L2 Cross-framework verified

This page aggregates 1 real-world runs of Qwen3.8-9B (IQ2_M) on NVIDIA RTX 4080 Super with vLLM, contributed by 1 independent source platforms; metrics are averages of published measurements.

181.18 tok/s

Decode

Decode speed

718.16 tok/s

Prefill

Prefill speed

1.97 s

TTFT

Time to first token

11.93 GB

VRAM

VRAM usage

L2 Cross-framework verified

1

Measured runs

1

Independent sources

Other

Source platforms

16 days ago

Last verified

Performance

  1. vLLM · IQ2_M (current)Decode 181.18 · Prefill 718.16 ·
  2. Strata · Q3_K_M Decode 213.25 · Prefill 900.93 ·
  3. Strata · Q3_K_S Decode 210.74 · Prefill 935.82 ·
  4. llama.cpp · IQ1_S Decode 204.16 · Prefill 1111.21 ·
  5. Ollama · IQ2_XS Decode 174.28 · Prefill 597.11 ·
  6. Ollama · Q3_K_M Decode 141.22 · Prefill 455.03 ·

Core figures

Decode (avg)
181.18 tok/s
Prefill (avg)
718.16 tok/s
TTFT (avg)
1.97 s
VRAM (avg)
11.93 GB
MTP acceptance rate
—
TTFB
1.45 GB
Power draw
214.6 W

Configuration

Member-level fields are taken from the most recent run

Model
Qwen3.8-9B
Quantization
IQ2_M
Framework
vLLM
Version
v0.32
Context length
4096 tokens
Batch size
16
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
NVIDIA RTX 4080 Super
Nominal VRAM
16 GB
Measured VRAM (avg)
11.93 GB
OS
Ubuntu 22.04
Driver
565.57
CUDA
12.6
Power draw
214.6 W

Sources & evidence

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

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

    v0.32 · Ubuntu 22.04 · CUDA 12.6 · 4096 ctx

    181.18 tok/s

    Decode

    718.16 tok/s

    Prefill

    1.97 s

    TTFT

    11.93 GB

    VRAM

    —

    MTP

    214.6 W

    Power draw

    L2 batch seed 2026-10-19: vLLM on hw_id=20 running model_id=52 at IQ2_M