Way to Run OpenClaw Locally on AMD Ryzen(amd.com)

Way to Run OpenClaw Locally on AMD Ryzen

https://www.amd.com/en/resources/articles/run-openclaw-locally-on-amd-ryzen-ai-max-and-radeon-gpus.html

1 comments

You can do same with Harbor in a couple of commands :)

# Pull Qwen 3.5 35B

harbor pull unsloth/Qwen3.5-35B-A3B-GGUF:UD-Q4_K_XL

# Set llama.cpp image to an optimised one for performance on Strix Halo

harbor config set llamacpp.image_rocm docker.io/kyuz0/amd-strix-halo-toolboxes:rocm-7.2

# Set default model for the claw

harbor config set openclaw.model unsloth/Qwen3.5-35B-A3B-GGUF:UD-Q4_K_XL

# Start both

harbor up openclaw llamacpp