Arao AI
Back to search results

gemma-4-12B-it · IQ2_M

Start a discussion
AMD Radeon RX 907016GBllama.cpp v0.38L2 Cross-framework verified

This page aggregates 1 real-world runs of gemma-4-12B-it (IQ2_M) on AMD Radeon RX 9070 with llama.cpp, contributed by 1 independent source platforms; metrics are averages of published measurements.

86.99 tok/s

Decode

Decode speed

431.57 tok/s

Prefill

Prefill speed

2.25 s

TTFT

Time to first token

15.2 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_XS Decode 114.22 · Prefill 439.54 ·
  2. llama.cpp · IQ2_M (current)Decode 86.99 · Prefill 431.57 ·

Core figures

Decode (avg)
86.99 tok/s
Prefill (avg)
431.57 tok/s
TTFT (avg)
2.25 s
VRAM (avg)
15.2 GB
MTP acceptance rate
—
TTFB
0.7 GB
Power draw
276.3 W

Configuration

Member-level fields are taken from the most recent run

Model
gemma-4-12B-it
Quantization
IQ2_M
Framework
llama.cpp
Version
v0.38
Context length
2048 tokens
Batch size
512
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
AMD Radeon RX 9070
Nominal VRAM
16 GB
Measured VRAM (avg)
15.2 GB
OS
Windows 11 23H2
Driver
amdgpu 6.11
CUDA
N/A
Power draw
276.3 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

    v0.38 · Windows 11 23H2 · CUDA N/A · 2048 ctx

    86.99 tok/s

    Decode

    431.57 tok/s

    Prefill

    2.25 s

    TTFT

    15.2 GB

    VRAM

    —

    MTP

    276.3 W

    Power draw

    L2 batch seed 2026-10-13: llama.cpp on hw_id=21 running model_id=65 at IQ2_M