Arao AI
Back to search results

Llama-3.1-8B-Instruct · IQ4_XS

Start a discussion
AMD Radeon RX 9070 XT16GBllama.cpp v0.38L2 Cross-framework verified

This page aggregates 1 real-world runs of Llama-3.1-8B-Instruct (IQ4_XS) on AMD Radeon RX 9070 XT with llama.cpp, contributed by 1 independent source platforms; metrics are averages of published measurements.

135.74 tok/s

Decode

Decode speed

591.43 tok/s

Prefill

Prefill speed

0.46 s

TTFT

Time to first token

14.53 GB

VRAM

VRAM usage

L2 Cross-framework verified

1

Measured runs

1

Independent sources

Other

Source platforms

Today

Last verified

Performance

  1. llama.cpp · IQ4_XS (current)Decode 135.74 · Prefill 591.43 ·
  2. vLLM · Q3_K_S Decode 162.87 · Prefill 574.19 ·
  3. vLLM · Q4_K_M Decode 144.21 · Prefill 558.74 ·

Core figures

Decode (avg)
135.74 tok/s
Prefill (avg)
591.43 tok/s
TTFT (avg)
0.46 s
VRAM (avg)
14.53 GB
MTP acceptance rate
—
TTFB
1.09 GB
Power draw
267.1 W

Configuration

Member-level fields are taken from the most recent run

Model
Llama-3.1-8B-Instruct
Quantization
IQ4_XS
Framework
llama.cpp
Version
v0.38
Context length
2048 tokens
Batch size
256
GPU layers
99
Flash Attention
Off

Hardware

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

GPU
AMD Radeon RX 9070 XT
Nominal VRAM
16 GB
Measured VRAM (avg)
14.53 GB
OS
Windows 11 23H2
Driver
amdgpu-pro 24.20
CUDA
12.6
Power draw
267.1 W

Sources & evidence

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

  1. L2 Cross-framework verifiedOtherOriginal link Verified on 2026-10-18

    v0.38 · Windows 11 23H2 · CUDA 12.6 · 2048 ctx

    135.74 tok/s

    Decode

    591.43 tok/s

    Prefill

    0.46 s

    TTFT

    14.53 GB

    VRAM

    —

    MTP

    267.1 W

    Power draw

    L2 batch seed 2026-10-18: llama.cpp on hw_id=16 running model_id=68 at IQ4_XS