Yes, AVX is specific to x86, though Apple’s chips have their own vector instructions.
NumPy wraps Intel’s x86-simd-sort C++ library.
Intel Publishes Blazing Fast AVX-512 Sorting Library, Numpy Switching To It For 10~17x Faster Sorts - Phoronix
Many if not all of those contributions were made by an Intel developer, who also wrapped at least one other Intel library to support AVX-512.
Intel Contributes AVX-512 Optimizations To Numpy, Yields Massive Speedups - Phoronix
Though it’s worth noting that this isn’t necessarily the fastest implementation.
sort-research-rs/writeup/intel_avx512/text.md at main · Voultapher/sort-research-rs · GitHub