Dandandan edited a comment on issue #1041: URL: https://github.com/apache/arrow-datafusion/issues/1041#issuecomment-925547561
> > For most queries I didn't see an impact in DataFusion using the simd feature. > > This is quite interesting, because SIMD is one of the big selling points of Databricks' photon engine. Do you know off the top of head whether this is expected? I would expect it to at least make a difference for aggregations? I am not sure whether it has a large improvement in the Photon engine on queries. I would expect the vectorization (vs row-based) and being implemented in efficiently, modern CPU-aware, written C++ rather than compiled to Java code making the largest impact, rather than the use of SIMD instructions. -- 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]
