Arao AI
Back to search results

DeepSeek-R1-Distill-Qwen-14B · IQ3_S

Start a discussion
AMD Radeon AI PRO R970032GBvLLM v0.20L2 Cross-framework verified

This page aggregates 1 real-world runs of DeepSeek-R1-Distill-Qwen-14B (IQ3_S) on AMD Radeon AI PRO R9700 with vLLM, contributed by 1 independent source platforms; metrics are averages of published measurements.

96.64 tok/s

Decode

Decode speed

451.11 tok/s

Prefill

Prefill speed

0.54 s

TTFT

Time to first token

21.5 GB

VRAM

VRAM usage

L2 Cross-framework verified

1

Measured runs

1

Independent sources

Other

Source platforms

5 days ago

Last verified

Performance

  1. vLLM · IQ3_S (current)Decode 96.64 · Prefill 451.11 ·
  2. vLLM · Q5_K_M Decode 71.18 · Prefill 263.48 ·
  3. Strata · IQ3_XS Decode 151.48 · Prefill 693.94 ·

Core figures

Decode (avg)
96.64 tok/s
Prefill (avg)
451.11 tok/s
TTFT (avg)
0.54 s
VRAM (avg)
21.5 GB
MTP acceptance rate
—
TTFB
0.77 GB
Power draw
308.8 W

Configuration

Member-level fields are taken from the most recent run

Model
DeepSeek-R1-Distill-Qwen-14B
Quantization
IQ3_S
Framework
vLLM
Version
v0.20
Context length
16384 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
AMD Radeon AI PRO R9700
Nominal VRAM
32 GB
Measured VRAM (avg)
21.5 GB
OS
Ubuntu 22.04
Driver
amdgpu 6.11
CUDA
N/A
Power draw
308.8 W

Sources & evidence

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

  1. L2 Cross-framework verifiedOtherOriginal link Verified on 2026-10-13

    v0.20 · Ubuntu 22.04 · CUDA N/A · 16384 ctx

    96.64 tok/s

    Decode

    451.11 tok/s

    Prefill

    0.54 s

    TTFT

    21.5 GB

    VRAM

    —

    MTP

    308.8 W

    Power draw

    L2 batch seed 2026-10-13: vLLM on hw_id=79 running model_id=71 at IQ3_S