Rich-T-kid commented on PR #11918: URL: https://github.com/apache/arrow-rs/pull/11918#issuecomment-6612234167
Ran `coalesce_kernels` and `filter_kernels` on an x86_64 Xeon with both the default target and `-C target-cpu=x86-64-v3`. Sparse coalescing is still 15–53% faster at 0.1–1% selectivity. With v3, the 10% and 80% cases regress by 9–26%, so I don’t think we should settle on this version yet. `pext` makes a big difference: the 50% Boolean filter drops from 10.6 µs to 1.58 µs with v3. -- 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]
