An index scan is always better $($fewer $1/0$ operations, faster run$-$time$)$ than a sequential scan, regardless of the processing model. Is this true? If yes, explain the reason. If not, give an example situation.