Any way to generate random numbers inside an @avx loop?

I am trying to parallellize a loop containing a random number generation, but the status of the rng is kept costant across each index of the loop, leading to a matrix with some rows.

Any idea how to solve this problem?

I could use [VectorizedRNG.jl](https://github.com/chriselrod/VectorizedRNG.jl) but I have no idea then how to glue it with the @avx loop.

Start by reproducing the random-number generation inside the @avx loop and read VectorizedRNG.jl to understand the proposed integration point. Done would mean the loop produces independent random values for each index rather than repeating the RNG state across rows, with a demonstrated working example or test.

Escrito por el modelo de indexación a partir del texto del issue.