New - Apple Neural Engine (ANE) backend for llama.cpp

This just showed up a couple of days ago on GitHub. Note that **ANE is the NPU in all Apple Silicon**, *not* the new 'Neural Accelerator' GPU cores that are only in M5. [(ggml-org/llama.cpp#10453)](https://github.com/ggml-org/llama.cpp/issues/10453#issuecomment-4148905254) \- Comment by **arozano