Hi Team, I had couple of scenarios in one application. 1) After stats collection on base tables, queries run faster , but with High Impact CPU for the insert statements. 2) There are some scenarios , where same as above or diffrent queries are running better with no stats and fixing the High Impact CPU after dropping stats. Looks ,Teradata optimizer is choosing a suboptimal query plan. Have we encountered any such scenarios and , what areas on code and objects , we should be looking at.