6 ms·

If you are using AVX, you might as well use AES-NI, which will probably be faster than PCG. Also, the xoshiro/xoroshiro RNGs are faster than PCG.

• isaachier 8y ago
  Not sure about that. The pcg C implementation uses some handrolled assembly and iirc already does SSE or AVX instructions.
  • whyever 8y ago
    Which implementation are you talking about? The only assembly I can see in the reference implementation is used because clang fails to optimize a rotation. There seems to be no explicit SIMD at all.
    • isaachier 8y ago
      My bad. They should add that.