The tremendous advances in communication technologies require next-generation communication systems to be optimally designed in terms of speed, reliability, cost, energy consumption, spectral efficiency, ecosystem compatibility, and sustainability. With these requirements in mind, this paper discusses downlink millimeter wave communication for a large-scale multiple-input multiple-output system equipped with a group of intelligent reconfigurable surfaces, one of the potential candidates for 6G wireless networks, considering 5G physical mmWave channel models. In addition to the precoder, beamformer, and combiner design of the proposed system  model, the theoretical average pairwise error probability and achievable rate expressions are derived. Furthermore, an upper bound for the achievable rate is derived. Simulation results clearly show that the performance of the proposed model is close to the theoretical bounds of the performance metrics and clearly demonstrates its superiority compared to a similar model based on index modulation.