Arao AI
Back to search results

Llama-3.1-8B-Instruct · IQ3_S

Start a discussion
NVIDIA RTX 3060 12GB12GBllama.cpp v0.33L2 Cross-framework verified

This page aggregates 1 real-world runs of Llama-3.1-8B-Instruct (IQ3_S) on NVIDIA RTX 3060 12GB with llama.cpp, contributed by 1 independent source platforms; metrics are averages of published measurements.

72.47 tok/s

Decode

Decode speed

300.81 tok/s

Prefill

Prefill speed

0.6 s

TTFT

Time to first token

11.4 GB

VRAM

VRAM usage

L2 Cross-framework verified

1

Measured runs

1

Independent sources

Other

Source platforms

17 days ago

Last verified

Performance

  1. llama.cpp · IQ3_S (current)Decode 72.47 · Prefill 300.81 ·
  2. vLLM · IQ2_XS Decode 94.63 · Prefill 430.36 ·
  3. vLLM · IQ2_M Decode 92.78 · Prefill 311.99 ·

Core figures

Decode (avg)
72.47 tok/s
Prefill (avg)
300.81 tok/s
TTFT (avg)
0.6 s
VRAM (avg)
11.4 GB
MTP acceptance rate
—
TTFB
0.58 GB
Power draw
158.7 W

Configuration

Member-level fields are taken from the most recent run

Model
Llama-3.1-8B-Instruct
Quantization
IQ3_S
Framework
llama.cpp
Version
v0.33
Context length
2048 tokens
Batch size
512
GPU layers
33
Flash Attention
On

Hardware

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

GPU
NVIDIA RTX 3060 12GB
Nominal VRAM
12 GB
Measured VRAM (avg)
11.4 GB
OS
Windows 11 23H2
Driver
565.57
CUDA
12.6
Power draw
158.7 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

    v0.33 · Windows 11 23H2 · CUDA 12.6 · 2048 ctx

    72.47 tok/s

    Decode

    300.81 tok/s

    Prefill

    0.6 s

    TTFT

    11.4 GB

    VRAM

    —

    MTP

    158.7 W

    Power draw

    L2 batch seed 2026-10-19: llama.cpp on hw_id=15 running model_id=68 at IQ3_S