It's accepted knowledge that the Google Swisstable is currently the fastest hashtable around. (The C++ implementation) But there are three variants, and this looks like the slowest of the three, but the best for dynamic languages.

It should depend a lot on what you put into it and the access pattern, no? You can talk about averages, about complexity, about benchmarks, but I feel like "fastest hashtable around" is a very odd expression, as if to desire the world to be a lot less complex than it is and for there to be exactly one best.

As I said there are two best, plus the worse third variant. Not implemented by google, only described, but now apparently implemented in Rust. Haven't checked closely, because Rust is kinda unreadable compared to C what exactly is going on. Need to see the SIMD assembly also. You cannot really trust much what's going on with Rust, as there's too much hype and lies, but it undeniably got better recently.

Is there anywhere we can read more about who says Google Swisstable is the fastest and why?