Single-header, 3D math library. Uses SIMDE as the backend, which allows for compilation on platforms that don't natively support SIMD. This means that the same code will run on x86-64, ARM, PowerPC, ...
A C++ compiler supporting C++20 (or later) CPU with the following SIMD support (at least one of them): AVX 2, AVX, SSE 4.2, SSE 4.1, SSSE3, SSE 3 Note: If you choose to enable MATH_LIBRARY_BUILD_TESTS ...
I used google to look for an arbitrary precision math library and found mapm and gnu mp. Which one of these two is the fastest? Is there another library faster than these two?