The information theoretic limits for the single user MIMO (Nt×Nr) communication channel are well established in literature for different scenarios of channel types and channel state information (CSI) availability. However, all obtained results are based on the linear system model in which the channel transfer function is defined as a single entity in which all environmental effects are captured. The capacity gain in MIMO over SISO systems is acquired by exploring the diversity in the channel transfer function. We show that, exploring the geographical location of the MIMO system nodes independently is of great influence to the overall system capacity. This is done by decomposing the channel transfer function into its equivalent transfer function in which we separate the effect of the signal bearing information (Angle of Arrival (AoA)) from the original channel. We show that, by exploring the geographical location of the system nodes, the capacity of the single user MIMO channel scales linearly with min {Nr2, Nt2} as opposed to min {Nr, Nt} if the channel is considered as a single transfer function. We show that these results holds for both deterministic, Rayleigh and Rician fading channels