rmuir commented on PR #13133: URL: https://github.com/apache/lucene/pull/13133#issuecomment-1965433852
> I'm surprised by how slow this is with AVX off given that this can be implemented with SSE2 :(. Yes, it is surprising: we found the same situation with VectorUtil byte[] methods. Seems hotspot just doesn't in fact implement/plumb the SSE2 for integer stuff (they do it for float though!). Also the "fallback" scalar code used by vector API is very very very slow. So you get things like 30x slowdowns in these situations: real performance trap. Traps that are easy to fall into with situations like virtualization where maybe AVX isn't plumbed to VMs. That's why we have so many checks in the code... -- 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] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
