Arao AI
Back to search results

DeepSeek-R1-Distill-Qwen-14B · Q5_K_S

Start a discussion
NVIDIA RTX PRO 6000 Blackwell 96GB96GBllama.cpp v0.43L2 Cross-framework verified

This page aggregates 1 real-world runs of DeepSeek-R1-Distill-Qwen-14B (Q5_K_S) on NVIDIA RTX PRO 6000 Blackwell 96GB with llama.cpp, contributed by 1 independent source platforms; metrics are averages of published measurements.

172.78 tok/s

Decode

Decode speed

845.2 tok/s

Prefill

Prefill speed

2.47 s

TTFT

Time to first token

28.56 GB

VRAM

VRAM usage

L2 Cross-framework verified

1

Measured runs

1

Independent sources

Other

Source platforms

16 days ago

Last verified

Performance

  1. llama.cpp · Q5_K_S (current)Decode 172.78 · Prefill 845.2 ·
  2. vLLM · IQ3_S Decode 236.15 · Prefill 1107.64 ·

Core figures

Decode (avg)
172.78 tok/s
Prefill (avg)
845.2 tok/s
TTFT (avg)
2.47 s
VRAM (avg)
28.56 GB
MTP acceptance rate
—
TTFB
1.38 GB
Power draw
338.6 W

Configuration

Member-level fields are taken from the most recent run

Model
DeepSeek-R1-Distill-Qwen-14B
Quantization
Q5_K_S
Framework
llama.cpp
Version
v0.43
Context length
8192 tokens
Batch size
256
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 RTX PRO 6000 Blackwell 96GB
Nominal VRAM
96 GB
Measured VRAM (avg)
28.56 GB
OS
Ubuntu 24.04
Driver
565.57
CUDA
12.6
Power draw
338.6 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

    v0.43 · Ubuntu 24.04 · CUDA 12.6 · 8192 ctx

    172.78 tok/s

    Decode

    845.2 tok/s

    Prefill

    2.47 s

    TTFT

    28.56 GB

    VRAM

    —

    MTP

    338.6 W

    Power draw

    L2 batch seed 2026-10-19: llama.cpp on hw_id=68 running model_id=71 at Q5_K_S