strix-point

Get internal audio (TAS2783 speakers + RT712 mic on AMD SoundWire) working on the HP OmniBook X Flip 14 (board 8EA1) under Linux

Local speech-to-text with AMD ROCm GPU acceleration

NixOS flake configuration for Lenovo ThinkPad with AMD Ryzen AI 7 PRO 350 (Strix Point) and Radeon 860M. Features modular design, AMD GPU optimizations, COSMIC desktop, and Home Manager integration.

Self-hosted AI image and video generation for AMD Ryzen AI APUs on Linux: Vulkan (Mesa RADV) via stable-diffusion.cpp, no ROCm. Web UI with queue, model manager, multi-user auth and system diagnostics.

⚡ Zero-dependency Homebrew tap for high-performance LLMs (CachyLLama, ROCmFPX, llama-ai). Optimized for AMD ROCm 7 (Strix Halo, RDNA3/4), Apple Silicon Metal, & Vulkan RADV.

Open, reproducible AMD XDNA1 + Strix Point XDNA2 compute on Linux: pinned builds, CPU-reference checks, persistent C/Python runners, and an LLM roadmap.

AMD NPU Enabler (aie2p) for Strix Point/Krackan (Zen 5, Ryzen AI 7 350) on Ubuntu 26.04. Validated 4.3 TOPS / 68us latency. Includes amdxdna v0.11.0 & XRT v2.23.0.

Fix for AMD XDNA NPU driver on Ryzen AI 300 (Strix Point/Halo) — SMU init bypass patch, systemd auto-loader, full root cause analysis

Running k2-fsa/OmniVoice voice cloning TTS on AMD Radeon 890M (gfx1150, Strix Point) integrated GPU — report, benchmark, and reproduction guide

🚀 Automated nightly builds & multi-backend releases (ROCm 7 TheRock, Vulkan RADV, CUDA, Metal) for CachyLLama & llama-ai on AMD Strix Halo (8060S), Strix Point (890M), Phoenix (780M), and Steam Deck with Homebrew support.

all-MiniLM-L6-v2 on the AMD Ryzen AI NPU (XDNA2), from hand-written MLIR-AIE kernels. C++ runtime, no Python, OpenAI-compatible /v1/embeddings. Faster than the CPU at half the cores and 1.9× less energy.