Arao AI
Back to search results

Llama-3.1-8B-Instruct · Q3_K_L

Start a discussion
AMD Radeon RX 907016GBStrata v0.51L2 Cross-framework verified

This page aggregates 1 real-world runs of Llama-3.1-8B-Instruct (Q3_K_L) on AMD Radeon RX 9070 with Strata, contributed by 1 independent source platforms; metrics are averages of published measurements.

194.02 tok/s

Decode

Decode speed

811.8 tok/s

Prefill

Prefill speed

1.63 s

TTFT

Time to first token

14.2 GB

VRAM

VRAM usage

L2 Cross-framework verified

1

Measured runs

1

Independent sources

Other

Source platforms

Today

Last verified

Performance

  1. vLLM · IQ1_S Decode 209.03 · Prefill 1072.51 ·
  2. vLLM · Q3_K_M Decode 141.43 · Prefill 703.16 ·
  3. Strata · Q3_K_L (current)Decode 194.02 · Prefill 811.8 ·

Core figures

Decode (avg)
194.02 tok/s
Prefill (avg)
811.8 tok/s
TTFT (avg)
1.63 s
VRAM (avg)
14.2 GB
MTP acceptance rate
0.61
TTFB
1.55 GB
Power draw
230.7 W

Configuration

Member-level fields are taken from the most recent run

Model
Llama-3.1-8B-Instruct
Quantization
Q3_K_L
Framework
Strata
Version
v0.51
Context length
4096 tokens
Batch size
256
GPU layers
12
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)
14.2 GB
OS
Ubuntu 24.04
Driver
amdgpu 6.11
CUDA
12.6
Power draw
230.7 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

    v0.51 · Ubuntu 24.04 · CUDA 12.6 · 4096 ctx

    194.02 tok/s

    Decode

    811.8 tok/s

    Prefill

    1.63 s

    TTFT

    14.2 GB

    VRAM

    0.61

    MTP

    230.7 W

    Power draw

    L2 batch seed 2026-10-18: Strata on hw_id=21 running model_id=68 at Q3_K_L