Series
2-part series running OpenClaw AI agent on the GMKtec EVO X2 (AMD Ryzen AI Max+ 395, 128GB unified memory) with SGLang backend and ROCm.
5 parts · 53 min of reading
Part 1: Tutorial
Set up a self-hosted OpenClaw AI agent on AMD Strix Halo: upgrade the kernel, install ROCm 7.2, and configure 128GB of unified memory for running local models.
Part 2: Tutorial
Wire OpenClaw into Uptime Kuma so a local AI agent catches outages and runs SSH diagnostics on its own, using a dispatcher, a host script, and restricted keys.
Part 3: Tutorial
Give your OpenClaw AI agent a real home. Self-host a Matrix server with Continuwuity in an LXC for persistent, searchable, threaded incident reports.
Part 4: Tutorial
Benchmark local LLMs on AMD Strix Halo's 128GB unified memory. Compare speed and answer quality across models to pick the best one for your OpenClaw agent.
Part 5: Tutorial
Back up and restore a self-hosted OpenClaw AI monitoring stack: what to capture, how to run the backup, and how to recover the whole setup after a disk failure.