Arao AI
Back to search results

Qwen 4 720B · IQ2_XXS

Start a discussion
NVIDIA DGX Spark128GBvLLM 0.7.3L0 Self-reported

This page aggregates 1 real-world runs of Qwen 4 720B (IQ2_XXS) on NVIDIA DGX Spark with vLLM, contributed by 1 independent source platforms; metrics are averages of published measurements.

1.1 tok/s

Decode

Decode speed

8.3 tok/s

Prefill

Prefill speed

12.05 s

TTFT

Time to first token

128 GB

VRAM

VRAM usage

L0 Self-reported

1

Measured runs

1

Independent sources

Other

Source platforms

Today

Last verified

Performance

  1. vLLM · IQ3_S Decode 0.8 · Prefill 6 ·
  2. vLLM · IQ2_XS Decode 1 · Prefill 7.5 ·
  3. vLLM · IQ2_XXS (current)Decode 1.1 · Prefill 8.3 ·

Core figures

Decode (avg)
1.1 tok/s
Prefill (avg)
8.3 tok/s
TTFT (avg)
12.05 s
VRAM (avg)
128 GB
MTP acceptance rate
—
TTFB
— GB
Power draw
— W

Configuration

Member-level fields are taken from the most recent run

Model
Qwen 4 720B
Quantization
IQ2_XXS
Framework
vLLM
Version
0.7.3
Context length
4096 tokens
Batch size
512
GPU layers
99
Flash Attention
On

Hardware

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

GPU
NVIDIA DGX Spark
Nominal VRAM
128 GB
Measured VRAM (avg)
128 GB
OS
Linux
Driver
570.65
CUDA
12.6

Sources & evidence

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

  1. L0 Self-reportedOtherOriginal link Verified on 2026-10-06

    0.7.3 · Linux · CUDA 12.6 · 4096 ctx

    1.1 tok/s

    Decode

    8.3 tok/s

    Prefill

    12.05 s

    TTFT

    128 GB

    VRAM

    —

    MTP

    — W

    Power draw

    L0 理论估算（add-500-records-batch2）