YouTube Series • Local AI Testing

PC Magik Lab

Local models, real GPUs

A series about running AI models on real, local GPUs. No cloud, no hype. Just measured, reproducible tests on a single rig.

RTX 3090

How we measure

Consistency first. Every test is designed to isolate one variable.

Control

Same prompt

Identical inputs across runs to eliminate prompt variance.

Control

Same rig

All tests run on the same hardware setup for fair comparison.

Method

One variable

Only one parameter changes per test. Everything else stays constant.

Method

Model reload

The model is reloaded before every run to ensure a clean state.

Open

Published data

Every prompt and result file is published for transparency and reproducibility.

Hardware

One rig. No swaps. Real measurements.

Test Rig

⚡
NVIDIA RTX 3090 24 GB
GPU used for all local inference tests
🧪
LM Studio
Model server for running and measuring local models

The same configuration is used across the series to keep results comparable.

3090

About

PC Magik Lab is a YouTube series focused on running AI models locally on real GPUs. The goal is clear testing, honest numbers, and practical insights for homelab builders who want local AI without cloud dependency.

Tests are reproducible, documented, and transparent. Every prompt and result file is published alongside the videos.

🧑‍💻

PC Magik

Mateusz Piekut
Computer service and homelab operator from Poland