If i take a long rod and calculate the MI about an axis running through the rod the MI when i take the origin as the center of the rod as origin is 1/6mr^2(i.e.1/12+1/12 with the held of perpendicular axis theorem). If i take the MI about the axis keeping the origin at the end of the rod the value becomes 1/3+1/3 i.e. 2/3. Isn't the MI about the axis through the cylinder supposed to be the same because i cannot imagine any difference in the motion of the rod. But there is supposed to be some difference? Where is it?