Awesome! 2 - 8X speed up! The performance improvement is impressive, thanks Weinan and Shujie who delivered this great feature.
With Vectorized Execution and Runtime filter feature finished, we have completed 2 of 3 features targets for HAWQ 2.4.0.0-incubating release: - [Done] New Feature: Pluggable Vectorized Execution Engine on HAWQ. - [Done] New Feature: Support Runtime Filter for HAWQ local hash join. - New Feature: Support accessing Hive table data by Pluggable Storage Framework. Regards, Radar On Tue, Jun 19, 2018 at 2:18 PM, Weinan Wang <wew...@pivotal.io> wrote: > Greetings, > > I am glad to announce that Vectorized Execution plugin has been completed > released in HAWQ. > > At present, it supports 7 types of data trigger the vectorized function > and reach two to eight times > acceleration comparing with regular execution in different performance > test queries. Furthermore, > it has an extended friendly framework, you can patch other types into as > your need. > > You can acquire the introduction document as follow: > https://cwiki.apache.org/confluence/display/HAWQ/Vectorized+Execution+ > Introduction <https://cwiki.apache.org/confluence/display/HAWQ/ > Vectorized+Execution+Introduction> > > > > > Regards, > > WANG Weinan
