Arao AI
Back to search results

Qwen3-8B · IQ3_XS

Start a discussion
AMD Radeon AI PRO R970032GBStrata v0.23L2 Cross-framework verified

This page aggregates 1 real-world runs of Qwen3-8B (IQ3_XS) on AMD Radeon AI PRO R9700 with Strata, contributed by 1 independent source platforms; metrics are averages of published measurements.

190.7 tok/s

Decode

Decode speed

730.63 tok/s

Prefill

Prefill speed

1.2 s

TTFT

Time to first token

12.52 GB

VRAM

VRAM usage

L2 Cross-framework verified

1

Measured runs

1

Independent sources

Other

Source platforms

5 days ago

Last verified

Performance

  1. llama.cpp · IQ2_M Decode 154.04 · Prefill 793.94 ·
  2. llama.cpp · Q3_K_S Decode 137.81 · Prefill 521.21 ·
  3. Strata · IQ1_S Decode 246.4 · Prefill 1052.77 ·
  4. Strata · IQ3_XS (current)Decode 190.7 · Prefill 730.63 ·

Core figures

Decode (avg)
190.7 tok/s
Prefill (avg)
730.63 tok/s
TTFT (avg)
1.2 s
VRAM (avg)
12.52 GB
MTP acceptance rate
0.59
TTFB
1.53 GB
Power draw
281.9 W

Configuration

Member-level fields are taken from the most recent run

Model
Qwen3-8B
Quantization
IQ3_XS
Framework
Strata
Version
v0.23
Context length
4096 tokens
Batch size
1024
GPU layers
5
Flash Attention
On

Hardware

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

GPU
AMD Radeon AI PRO R9700
Nominal VRAM
32 GB
Measured VRAM (avg)
12.52 GB
OS
macOS 15.0
Driver
amdgpu 6.10
CUDA
N/A
Power draw
281.9 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

    v0.23 · macOS 15.0 · CUDA N/A · 4096 ctx

    190.7 tok/s

    Decode

    730.63 tok/s

    Prefill

    1.2 s

    TTFT

    12.52 GB

    VRAM

    0.59

    MTP

    281.9 W

    Power draw

    L2 batch seed 2026-10-13: Strata on hw_id=79 running model_id=53 at IQ3_XS