The implementation difference for this mode from round to highest magnitude is that the sign bit is used directly as the carry-in (see This Figure ). There is no direct MATLAB software equivalent of this operation. One minor advantage of rounding towards zero is that it does not cause overflow situations.