ozgrakkurt commented on PR #4441: URL: https://github.com/apache/arrow-rs/pull/4441#issuecomment-1609930665
> If autovectorization / avx is of concern we could also bring back feature detection using the `multiversion` crate. > > I know also of a few (cloud) users are enabling a specific cpu target, e.g. by setting `target-cpu=skylake` etc as the instance type / cpu is often known in advance. With `target-cpu=native` on avx2 enabled cpu, library is 3x as fast on `insert` and 1.4x as fast on `contains` also library code is very easy to use without any requirement on cpu targets or anything like that -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
