Arao AI
Back to search results

Llama-3-8B-Instruct · IQ3_S

Start a discussion
Apple M2 UltraOllama v0.50L2 Cross-framework verified

This page aggregates 1 real-world runs of Llama-3-8B-Instruct (IQ3_S) on Apple M2 Ultra with Ollama, contributed by 1 independent source platforms; metrics are averages of published measurements.

116.94 tok/s

Decode

Decode speed

440.9 tok/s

Prefill

Prefill speed

0.47 s

TTFT

Time to first token

13.19 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 · IQ2_M Decode 144.03 · Prefill 634.04 ·
  2. llama.cpp · Q2_K Decode 151.53 · Prefill 649.5 ·
  3. Ollama · IQ3_S (current)Decode 116.94 · Prefill 440.9 ·
  4. Ollama · Q3_K_M Decode 114.21 · Prefill 572.36 ·

Core figures

Decode (avg)
116.94 tok/s
Prefill (avg)
440.9 tok/s
TTFT (avg)
0.47 s
VRAM (avg)
13.19 GB
MTP acceptance rate
—
TTFB
0.94 GB
Power draw
118.6 W

Configuration

Member-level fields are taken from the most recent run

Model
Llama-3-8B-Instruct
Quantization
IQ3_S
Framework
Ollama
Version
v0.50
Context length
4096 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
Apple M2 Ultra
Measured VRAM (avg)
13.19 GB
OS
macOS 26.1
Power draw
118.6 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

    v0.50 · macOS 26.1 · 4096 ctx

    116.94 tok/s

    Decode

    440.9 tok/s

    Prefill

    0.47 s

    TTFT

    13.19 GB

    VRAM

    —

    MTP

    118.6 W

    Power draw

    L2 batch seed 2026-10-19: Ollama on hw_id=103 running model_id=69 at IQ3_S