AMD Strix Halo NPU becomes usable for local LLM inference

A Reddit user reports that the NPU on AMD Ryzen AI Max+ (Strix Halo) processors is now functional for AI inference, after relying solely on GGUF and Vulkan for the past year. The post highlights progress from the AMD Ryzen AI team in enabling NPU support for local LLM workloads.