Jeff Pflueger

AI experiments part 1: Setting Up A Local AI Workstation

October 1, 2025

I am a continual student. Learning new ways to work is important and fun. Its 2025 and I want to catch up quickly on what I can do to super size my productivity with my coding background and available AI tools.

To start I want a local desktop environment to build and experiment.

Step 1 - Build Local AI Development Environment

* For a local AI playground got a GMKtec mini PC - the EVO-X1. Features the AMD Ryzen Al 9 HX 370 as a CPU and the Radeon 890M GPU. Won't be a powerhouse, but it is inexpensive and powerful enough at this stage for my purposes.

* Installed Ubuntu 24.04 as dual boot (so I can also still use Windows) - intention is to use Ubuntu operating system for my AI work and keep my activities firewalled from other workstations.

* Installed Docker and Docker image released from AMD for Pytorch with ROCm support - required to leverage the GPU for faster processing