Arao AI
Back to search results

elephant-1 · IQ2_XS

Start a discussion
NVIDIA H2097GBvLLM v0.49L2 Cross-framework verified

This page aggregates 1 real-world runs of elephant-1 (IQ2_XS) on NVIDIA H20 with vLLM, contributed by 1 independent source platforms; metrics are averages of published measurements.

68.56 tok/s

Decode

Decode speed

374.49 tok/s

Prefill

Prefill speed

2.17 s

TTFT

Time to first token

11.17 GB

VRAM

VRAM usage

L2 Cross-framework verified

1

Measured runs

1

Independent sources

Other

Source platforms

13 days ago

Last verified

Performance

  1. vLLM · FP16 Decode 25.76 · Prefill 119.08 ·
  2. vLLM · IQ2_XS (current)Decode 68.56 · Prefill 374.49 ·
  3. Ollama · Q5_K_M Decode 36.62 · Prefill 154.85 ·

Core figures

Decode (avg)
68.56 tok/s
Prefill (avg)
374.49 tok/s
TTFT (avg)
2.17 s
VRAM (avg)
11.17 GB
MTP acceptance rate
—
TTFB
0.55 GB
Power draw
384.5 W

Configuration

Member-level fields are taken from the most recent run

Model
elephant-1
Quantization
IQ2_XS
Framework
vLLM
Version
v0.49
Context length
32768 tokens
Batch size
32
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 H20
Nominal VRAM
97 GB
Measured VRAM (avg)
11.17 GB
OS
Ubuntu 22.04
Driver
572.83
CUDA
12.6
Power draw
384.5 W

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 2026-10-22

    v0.49 · Ubuntu 22.04 · CUDA 12.6 · 32768 ctx

    68.56 tok/s

    Decode

    374.49 tok/s

    Prefill

    2.17 s

    TTFT

    11.17 GB

    VRAM

    —

    MTP

    384.5 W

    Power draw

    L2 batch seed 2026-10-22: vLLM on hw_id=56 running model_id=95 at IQ2_XS