Arao AI
Back to search results

Llama-3-8B-Instruct · Q8_0

Start a discussion
NVIDIA RTX 5070 SUPER18GBllama.cpp b4200L0 Self-reported

This page aggregates 1 real-world runs of Llama-3-8B-Instruct (Q8_0) on NVIDIA RTX 5070 SUPER with llama.cpp, contributed by 1 independent source platforms; metrics are averages of published measurements.

81.3 tok/s

Decode

Decode speed

609.8 tok/s

Prefill

Prefill speed

0.16 s

TTFT

Time to first token

9.1 GB

VRAM

VRAM usage

L0 Self-reported

1

Measured runs

1

Independent sources

Other

Source platforms

Today

Last verified

Performance

  1. llama.cpp · Q8_0 (current)Decode 81.3 · Prefill 609.8 ·
  2. llama.cpp · Q5_K_M Decode 118.5 · Prefill 888.8 ·
  3. llama.cpp · Q4_K_M Decode 142.6 · Prefill 1069.5 ·

Core figures

Decode (avg)
81.3 tok/s
Prefill (avg)
609.8 tok/s
TTFT (avg)
0.16 s
VRAM (avg)
9.1 GB
MTP acceptance rate
—
TTFB
— GB
Power draw
— W

Configuration

Member-level fields are taken from the most recent run

Model
Llama-3-8B-Instruct
Quantization
Q8_0
Framework
llama.cpp
Version
b4200
Context length
4096 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
NVIDIA RTX 5070 SUPER
Nominal VRAM
18 GB
Measured VRAM (avg)
9.1 GB
OS
Linux
Driver
570.65
CUDA
12.6

Sources & evidence

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

  1. L0 Self-reportedOtherOriginal link Verified on 2026-10-06

    b4200 · Linux · CUDA 12.6 · 4096 ctx

    81.3 tok/s

    Decode

    609.8 tok/s

    Prefill

    0.16 s

    TTFT

    9.1 GB

    VRAM

    —

    MTP

    — W

    Power draw

    L0 理论估算（add-1000-records-batch3）