Ultra-wideband reception of signals is desired in several applications such as spectrum detection and weather radar [1]. Yet, the design of the wideband radio-frequency front-end is challenging due to the limitation of the Bode-Fano theory [2]. Distributed amplifier (DA) features large fractional bandwidth (BW) by merging the device parasitic capacitance into the artificial transmission lines (AT-lines). DA is almost the unique circuit topology to obtain more than one-octave BW in the millimeter-wave (mm-wave) band. As the frequency increases to the mm-wave band, CMOS DAs suffer from limited gain, poor gain flatness, and high noise Figure (NF). Meanwhile, most of the prior DAs are single-ended (SE) [3], but the differential outputs are preferred due to the benefits of the interference cancellation and even-order distortion suppression. Passive baluns [4] are unable to provide sufficient fractional BW for CMOS DAs. Active baluns [5-6] are considerable, but it is hard to achieve good performances, such as gain, linearity, and differential balance at the mm-wave band.