Historically, power systems were dispatched in a centralized way when all control actions are processed from one control center. However, centralized optimization is not applicable for large interconnected power systems of several independent Transmission System Operators (TSOs) due to the enormous size of the optimization problem and the tendency of each TSO to preserve the autonomy of own controlled area and secure primary data. Alternatively, there are two approaches to solve the problem: decentralized and distributed. This paper presents the advantages and drawbacks of methods in each group as well as numerical results of subgradient and interior point methods performance. The distributed optimization by subgradient method is carried out with the use of Python and RastrWin, which is the main program of System Operator in Russia. The decentralized optimization by interior point method is more appropriate to Super Grid projects than the subgradient method due to faster convergence and no need of the central coordinator. Both methods are implemented on 14 and 118 bus IEEE systems and mutually compared by the objective function value, the number of iterations and computation time.