rx580-local-ai-guide
A technical guide documenting how to run local LLM and image-generation inference on an AMD RX 580 GPU through Vulkan, after AMD dropped ROCm support for that hardware generation.
This repository is a guide rather than a packaged tool: it records how to get modern local AI inference running on an AMD RX 580 GPU (Polaris/GCN4), a card AMD's official ROCm no longer supports. It explains compiling llama.cpp and stable-diffusion.cpp from source with Vulkan support so the GPU can still run large language model and image-generation inference.
The guide covers hardware and driver requirements, benchmarks for LLM inference and image generation on this setup, quick-start instructions for both workloads, integration with OpenWebUI through Docker, and additional recipes for audio transcription with whisper.cpp, voice cloning with Applio RVC, and video generation with AnimateDiff, along with notes on approaches that failed.