Massive MIMO (multiple input multiple output) systems are best suitable for mmWave communications improving throughput and spectral efficiency in 5G. Beamforming is a wireless technique adopted by massive MIMO and used in 4G & 5G to increase the directivity and energy efficiency by focusing the signal in specific direction supporting only single user transmission with one data stream. Precoding is a generalized form of beamforming to support multi-stream transmission in multi-antenna wireless communication, such that combination of analog and digital precoding forms the hybrid precoding which shows good performance with less complexity. Successive Interference Cancelation (SIC) is a technique in which optimization of different antenna arrays will be done one by one such that while optimizing the capacity of specific array contribution of earlier optimized array is removed from the total capacity and precoder of that specific optimizing array will be computed. In this paper we have designed the SIC based hybrid precoding using sub-connected and fully-connected structures for multi-user (MU) case and compared them with the optimal precoding for mmWave massive MIMO systems in a 3D scenario, where both azimuth and elevation angles are taken into account in the channel. The proposed algorithms are simulated in the MATLAB and compared their performance with different parameters and shown that SIC-based scheme is near-optimal for multi-user case.