Here is an example of Rust implementation being faster than C.
Discussions on HN of that post at the time more or less debunked the title's claim, from what I recall.
Hat tip! Here is the original HN discussion: https://news.ycombinator.com/item?id=43381512
Another thing about that library, like so many "Rust wins" that we are seeing: There is loads of unsafe code.
Yeah that was the one. I too left it unconvinced that Rust can get measurably faster than C, for what it's worth. But as said in another comment, even being 1-2% around the same ballpark is a huge win already and it would be enough for me.