Arao AI
Back to search results

Qwen3.6-35B-A3B · Q3_K_M

Start a discussion
NVIDIA H100 80GB HBM380GBvLLM v0.23L2 Cross-framework verified

This page aggregates 1 real-world runs of Qwen3.6-35B-A3B (Q3_K_M) on NVIDIA H100 80GB HBM3 with vLLM, contributed by 1 independent source platforms; metrics are averages of published measurements.

355.61 tok/s

Decode

Decode speed

1,722.65 tok/s

Prefill

Prefill speed

0.33 s

TTFT

Time to first token

52.08 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 · IQ4_XS Decode 288.72 · Prefill 1282.53 ·
  2. vLLM · Q3_K_M (current)Decode 355.61 · Prefill 1722.65 ·
  3. Ollama · IQ2_M Decode 321.92 · Prefill 1762.51 ·
  4. Ollama · Q8_0 Decode 296.95 · Prefill 1140.84 ·

Core figures

Decode (avg)
355.61 tok/s
Prefill (avg)
1,722.65 tok/s
TTFT (avg)
0.33 s
VRAM (avg)
52.08 GB
MTP acceptance rate
—
TTFB
2.84 GB
Power draw
611.4 W

Configuration

Member-level fields are taken from the most recent run

Model
Qwen3.6-35B-A3B
Quantization
Q3_K_M
Framework
vLLM
Version
v0.23
Context length
32768 tokens
Batch size
8
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 H100 80GB HBM3
Nominal VRAM
80 GB
Measured VRAM (avg)
52.08 GB
OS
Ubuntu 22.04
Driver
572.83
CUDA
12.8
Power draw
611.4 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

    v0.23 · Ubuntu 22.04 · CUDA 12.8 · 32768 ctx

    355.61 tok/s

    Decode

    1,722.65 tok/s

    Prefill

    0.33 s

    TTFT

    52.08 GB

    VRAM

    —

    MTP

    611.4 W

    Power draw

    L2 batch seed 2026-10-19: vLLM on hw_id=55 running model_id=47 at Q3_K_M