Apple Neural Engine (ANE) LLM inference engine — reverse-engineered private APIs, Metal GPU shaders, hybrid ANE+GPU+CPU on Apple Silicon. 32 tok/s matching llama.cpp, 3.6 TFLOPS fused ANE mega-kernels.
macos rust reverse-engineering quantization ane npu edge-ai deltanet on-device-ai neural-engine apple-silicon apple-neural-engine llm-inference qwen gguf metal-gpu
-
Updated
Mar 5, 2026 - Rust