[oconnor663]

The part about data dependencies across loop iterations is fascinating to me, becuase it's mostly invisible even when you look at the generated assembly. There's a related optimization that comes up in implementations of ChaCha/BLAKE, where we permute columns around in a kind of weird order, because it breaks a data dependency for an operation that's about to happen: https://github.com/sneves/blake2-avx2/pull/4#issuecomment-50...

[ummonk]

The pipelining issue is interesting because my reaction becomes “shouldn’t the CPU just come with a larger vector size and then operate on chunks within the vector to optimize pipelining?” but then I realize I’m just describing a GPU.