At least on x86, the fastest intrinsics for shuffling the contents of a vector or blending data from two vectors take an immediate operand, which must be a compile-time constant. So there would be a use case for a compile-time version of xsimd::select(), as it could use these faster instructions.

An example of prior art for this is the shuffle() instruction family in bSIMD:

• https://develop

There are hardware instructions on X86 for approximate inverse square root and square root which can improved upon via newton's method.
See the for loop at the bottom of bodies_advance here: https://benchmarksgame-team.pages.debian.net/benchmarksgame/program/nbody-rust-7.html
It would look a lot better if this example could be re-written to use purely safe, normal looking rust code.