Radeon - Tag - Lorenzo's Bloghttps://www.k8s.it/tags/radeon/Radeon - Tag - Lorenzo's BlogHugo -- gohugo.ioenSun, 07 Jun 2026 00:00:00 +0000Running a Local LLM on AMD Radeon 780M — gfx1103, ROCm, and the GPU That Wasn't Supposed to Workhttps://www.k8s.it/posts/running-a-local-llm-on-amd-radeon-780m-gfx1103-rocm-and-the-gpu-that-wasnt-supposed-to-work/Sun, 07 Jun 2026 00:00:00 +0000Lorenzo Girardihttps://www.k8s.it/posts/running-a-local-llm-on-amd-radeon-780m-gfx1103-rocm-and-the-gpu-that-wasnt-supposed-to-work/

Table of Contents

  • The Machine
  • The Problem: gfx1103 Doesn’t Exist
  • GTT Memory — 24 GB for Free
  • The ROCm Stack
  • Getting GPU Inference Working
  • Optimizing: The Hidden GPU Clock Problem
  • Benchmarks — Every Configuration Tested
  • The Surprising Finding: CPU Beats GPU on Generation
  • Monitoring with Collectd and Grafana
  • What the Dashboard Actually Shows
  • Lessons Learned

I wanted a local AI box. Not a cloud API with latency and per-token billing. Not a GPU workstation that sounds like a jet engine. A quiet mini-PC that runs a capable model at home, on my desk, forever, for free.

]]>