AMD Strix Halo RDMA Cluster Setup Guide

This guide provides instructions for setting up a two-node AMD Strix Halo cluster using Intel E810 RDMA networking to enable distributed AI model inference through vLLM with Tensor Parallelism. The setup leverages RoCE v2 protocol to achieve ultra-low latency (~5µs) communication between nodes, allowing large language models to be split across the two systems for faster processing. The guide covers hardware prerequisites, Fedora host configuration, RDMA verification, and cluster deployment procedures, with an automated toolbox script that simplifies installation and network configuration.