The data flow is two simultaneous full scans for employees and departments. Then a hash join and finally a GROUP BY sort.