Teradata DISTINCT VS GROUP BY

Which is better DISTINCT or GROUP BY in Teradata?

     Till Teradata 12, we all knew that DISTINCT uses more spool since it picks the each row from ever amp and redistributes them to appropriate AMP then SORT the data to find the duplicates.(So the output has sorted output)

      Whereas GROUP BY uses "Aggregate" concept (AMP local grouping) to remove duplicates so consumes less spool(Output won’t be sorted, you need to add ORDER BY incase of sorted output)

      From Teradata 13/14, All DISTINCT will automatically rewritten to GROUP BY and apply some new algorithms developed by Teradata to get faster response… 

You can run a simple query for distinct and group by. You can observe they have same explain plan from TD 13

Explain SEL Distinct COL1 FROM DB.TABLE1; -- No Sorted output now

Explain SEL COL1 FROM DB.TABLE1 GROUP BY 1; -- Same explain as above

So use DISTINCT and GROUP BY for its created purpose, i.e. Distinct to find unique records and group by for analytic functions to give an aggregate result. don’t worry about performance or spool space issues any more. 

Happy Reading and blog is open for any queries/Clarification!!