Arao AI
Back to search results

gemma-4-E2B-it · IQ4_XS

Start a discussion
AMD Strix Point (Radeon 890M)llama.cpp v0.28L2 Cross-framework verified

This page aggregates 1 real-world runs of gemma-4-E2B-it (IQ4_XS) on AMD Strix Point (Radeon 890M) with llama.cpp, contributed by 1 independent source platforms; metrics are averages of published measurements.

106.04 tok/s

Decode

Decode speed

563.55 tok/s

Prefill

Prefill speed

0.58 s

TTFT

Time to first token

5.21 GB

VRAM

VRAM usage

L2 Cross-framework verified

1

Measured runs

1

Independent sources

Other

Source platforms

2 days ago

Last verified

Performance

  1. llama.cpp · IQ1_S Decode 145.7 · Prefill 692.08 ·
  2. llama.cpp · IQ4_XS (current)Decode 106.04 · Prefill 563.55 ·
  3. Ollama · FP16 Decode 47.28 · Prefill 250.08 ·
  4. Ollama · Q8_0 Decode 57.25 · Prefill 264.02 ·

Core figures

Decode (avg)
106.04 tok/s
Prefill (avg)
563.55 tok/s
TTFT (avg)
0.58 s
VRAM (avg)
5.21 GB
MTP acceptance rate
—
TTFB
0.85 GB
Power draw
65 W

Configuration

Member-level fields are taken from the most recent run

Model
gemma-4-E2B-it
Quantization
IQ4_XS
Framework
llama.cpp
Version
v0.28
Context length
16384 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 Strix Point (Radeon 890M)
Measured VRAM (avg)
5.21 GB
OS
Ubuntu 22.04
Driver
570.86
CUDA
N/A
Power draw
65 W

Sources & evidence

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

  1. L2 Cross-framework verifiedOtherOriginal link Verified on 2026-10-16

    v0.28 · Ubuntu 22.04 · CUDA N/A · 16384 ctx

    106.04 tok/s

    Decode

    563.55 tok/s

    Prefill

    0.58 s

    TTFT

    5.21 GB

    VRAM

    —

    MTP

    65 W

    Power draw

    L2 batch seed 2026-10-16: llama.cpp on hw_id=89 running model_id=66 at IQ4_XS