SIMD-accelerated vector similarity kernels, SimSIMD 6.x API
SimSIMD provides SIMD-accelerated distance and similarity kernels, such as cosine, inner product and Euclidean distances, for mixed-precision vectors on Arm and x86, with runtime dispatch to the best instruction set. Version 7 continues as NumKong with a different API.
SimSIMD provides SIMD-accelerated distance and similarity kernels, such as cosine, inner product and Euclidean distances, for mixed-precision vectors on Arm and x86, with runtime dispatch to the best instruction set. Version 7 continues as NumKong with a different API.
To install simsimd, run the following command in macOS terminal (Applications->Utilities->Terminal)
sudo port install simsimd
To see what files were installed by simsimd, run:
port contents simsimd
To later upgrade simsimd, run:
sudo port selfupdate && sudo port upgrade simsimd
Reporting an issue on MacPorts Trac
The MacPorts Project uses a system called Trac to file tickets to report bugs and enhancement requests.
Though anyone may search Trac for tickets, you must have a GitHub account in order to login to Trac to create tickets.