ROCm

ROCm is AMD's GPU-compute stack — its answer to NVIDIA's CUDA — and it's the software layer that decides how well the Radeon 8060S iGPU in a Strix Halo box accelerates AI work. On the Ryzen AI Max+ 395 boxes, ROCm/Linux GPU-compute is still rough compared with CUDA's maturity, which is the recurring caveat on AMD hardware: the memory capacity is there, but the software path is less polished. Many users route around it via llama.cpp's Vulkan backend instead. It's improving quickly but isn't yet CUDA-smooth.

ROCm maturity is the hidden cost of the cheaper AMD capacity boxes — budget some setup friction versus a plug-and-play NVIDIA or Apple path. If you want zero-fuss tooling, that friction is a real reason to pay up for GB10 or Apple.