Show HN: Hsdlib – A C Library for Vector Similarity with SIMD Acceleration

Hsdlib is a small C library for fast vector distance and similarity calculations.

At the moment, it supports:

    - Euclidean, Manhattan, and Hamming distances

    - Dot product, cosine, and Jaccard similarities

Hsdlib uses SIMD acceleration (AVX, AVX2, AVX512, NEON, and SVE instructions) to speed things up.

GitHub link: https://github.com/habedi/hsdlib