AI & Compute

AMD Strix Halo Local AI Compatibility

Which local-AI runtimes work on Strix Halo (Ryzen AI Max+ / Radeon 8060S) — llama.cpp on ROCm and Vulkan tested at pinned versions, PyTorch tracked but untested, with backend and status per combination.

Updated August 23, 2026

What this page answers: which local-AI runtimes work on Strix Halo (Ryzen AI Max+ / Radeon 8060S), at which pinned version, and how each row was established.

Overall matrix:

Runtime / frameworkStatusBackendTested versionNotes
llama.cppTested (2026-08-24)ROCm/HIPc060ca974 (build 10603) · ROCm 7.14benchmark · setup guide
llama.cppTested (2026-08-24)Vulkan (RADV)c060ca974 (build 10603) · Mesa 26.1.7benchmark · setup guide
PyTorchUnverified (checked 2026-08-24)——no runs made on this machine

Scope: inference servers (vLLM, SGLang) are not evaluated here; they will be added to the matrix only once actually attempted on this hardware.

llama.cpp

ROCm / HIP backend

Status: Tested (2026-08-24) · llama.cpp c060ca974 (build 10603), -DGGML_HIP=ON, ROCm 7.14 · Setup: Install ROCm 7.14 · Results: ROCm vs Vulkan benchmark

Vulkan backend

Status: Tested (2026-08-24) · llama.cpp c060ca974 (build 10603), -DGGML_VULKAN=ON, Mesa 26.1.7 (RADV, kisak-mesa PPA) · Setup: Vulkan compute guide · Results: benchmark

PyTorch

ROCm wheel / source build for gfx1151

Status: Unverified (checked 2026-08-24) · No PyTorch runs are recorded on this machine. To test it properly: pin the version, record exactly how it was installed for gfx1151 (ROCm wheel index, env vars such as HSA_OVERRIDE_GFX_VERSION if used, or a source build), and run a two-line GPU tensor check (device available + one matmul on-device) before upgrading this row to Tested.

What’s actually installed on this box (as captured)

  • llama.cpp c060ca974 (build 10603) in two builds from the same commit — HIP and Vulkan
  • ROCm 7.14, Mesa 26.1.7 (kisak-mesa PPA, RADV)
  • The box also carries libnvidia-gl-595 despite being all-AMD: its Vulkan ICD is present but non-functional (no NVIDIA GPU), and the loader skips it harmlessly — a headless vulkaninfo run showed ERROR ... libGLX_nvidia.so.0: Could not get 'vkCreateInstance' followed by Skipping this driver. Where that package came from is not recorded here, and nothing on this page depends on it.

How to verify a combination (recipe)

  1. Install the runtime at a pinned version.
  2. Run the smallest real workload (not just an import check).
  3. Record: version string, exact commands, output proof, memory usage.
  4. Update the matrix row; keep “Tested” tied to that version — older entries stay dated rather than silently upgraded.

Sources and upstream documentation