#ripgrep thought I'd found the holy grail, something that would either use "all cores" or "the full throughput of the disk" but...I seem to see neither? :| slightly faster than ag though, thanks!
ripgrep will use parallelism by default, and if you're searching a corpus that isn't in your operating system's I/O cache, then ripgrep will indeed be disk bound. If you have more data that you can share (as much as possible), then please file an issue:https://github.com/BurntSushi/ripgrep/issues …