Dandandan commented on PR #8854: URL: https://github.com/apache/arrow-rs/pull/8854#issuecomment-3551432757
> The benchmarks show a slowdown for some operations for some reason > > > buffer_binary_ops/and_with_offset 1.13 1486.2±3.20ns 9.6 GB/sec 1.00 1320.5±3.78ns 10.8 GB/sec > > However, given the duration of the benchmark, I am thinking maybe this is cache lines or something. > > I have an idea of how to improve the benchmarks so they are less noisy (basically run them in a 100x loop) Might also because of the allocation? Looks like `and_with_offset` and `and` are not a over a power of two inputs. -- 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]
