What it does: Our CLI aggregates AMD hardware availability from multiple providers in one place, showing pricing, specs, and real-time availability for both AMD CPUs (EPYC, Ryzen) and GPUs (MI series, Radeon). You can search, compare, and get direct links to provision instances without hunting through individual provider dashboards.
Why we built this: We were at the AGI House hackathon (sponsored by AMD) trying to deploy a model to AMD compute. We discovered the ecosystem is surprisingly fragmented - there's no easy way to find where AMD hardware is actually available or compare options across providers.
The pain was real: our organizer Kyle spent hours on calls with AMD's partnership team trying to get API access. That failed, no one was available to help us, we had to wait for procurement for API keys and they asked us to download a VPN and jump through multiple hoops just to access certain providers' platforms. This shouldn't be so hard.
How to try it:
chmod +x camd.py
sudo cp camd.py /usr/local/bin/camd
camd list
No signup required. The tool queries public APIs and availability endpoints.What's different: Unlike provider-specific tools, this gives you a unified view of the entire AMD compute landscape - from EPYC servers to MI300X accelerators. Think of it as "whereami" but for AMD hardware.
GitHub: https://github.com/modelturnedgeek/CheaperNvidia
We built this in 24 hours, so it's rough around the edges. Would love feedback on what providers to add next and what features would make this actually useful for your workflows.