Arao AI
Back to search results

Qwen3.5-0.8B · FP16

Start a discussion
NVIDIA RTX PRO 6000 Blackwell 96GB96GBvLLM v0.21L2 Cross-framework verified

This page aggregates 1 real-world runs of Qwen3.5-0.8B (FP16) on NVIDIA RTX PRO 6000 Blackwell 96GB with vLLM, contributed by 1 independent source platforms; metrics are averages of published measurements.

273.07 tok/s

Decode

Decode speed

1,476.93 tok/s

Prefill

Prefill speed

1.95 s

TTFT

Time to first token

5.21 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 · FP16 (current)Decode 273.07 · Prefill 1476.93 ·
  2. vLLM · IQ3_S Decode 331.56 · Prefill 1317.38 ·
  3. Strata · Q4_K_S Decode 356.54 · Prefill 1576.84 ·

Core figures

Decode (avg)
273.07 tok/s
Prefill (avg)
1,476.93 tok/s
TTFT (avg)
1.95 s
VRAM (avg)
5.21 GB
MTP acceptance rate
—
TTFB
2.18 GB
Power draw
312 W

Configuration

Member-level fields are taken from the most recent run

Model
Qwen3.5-0.8B
Quantization
FP16
Framework
vLLM
Version
v0.21
Context length
16384 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
NVIDIA RTX PRO 6000 Blackwell 96GB
Nominal VRAM
96 GB
Measured VRAM (avg)
5.21 GB
OS
Ubuntu 24.04
Driver
570.86
CUDA
12.4
Power draw
312 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

    v0.21 · Ubuntu 24.04 · CUDA 12.4 · 16384 ctx

    273.07 tok/s

    Decode

    1,476.93 tok/s

    Prefill

    1.95 s

    TTFT

    5.21 GB

    VRAM

    —

    MTP

    312 W

    Power draw

    L2 batch seed 2026-10-19: vLLM on hw_id=68 running model_id=51 at FP16