In Teradata, is there any difference in CPU usage between “select distinct value” and “select value group by value”?