Abstract <p>Simulation of traction power-supply systems is performed for large electrified railway polygons: mainline electric railways or entire lines of metro systems. The capabilities of computer technology and theoretical developments make the calculation of quasi-steady-state modes with acceptable waiting times possible; however, the range of tasks for software complexes used in traction power-supply system calculations is expanding. There is a need to calculate transient processes and electromagnetic compatibility, take into account the actual voltage at the current collector of the electric rolling stock, analyze emergency modes including solid and arc short circuits, etc. This will enable solving problems related to selecting appropriate settings for protective devices, assessing the impact of generator operating modes of electric rolling stock on the breaking capacity of high-speed circuit breakers, ensuring the safety of electric locomotive passage through insulating joints under voltage, calculating the required reactor power and parameters of filter devices, and so on. Under such conditions, calculation again becomes a labor-intensive computational task. Opportunities to increase calculation speed lie in the use of parallel computing, where acceleration occurs both by performing several calculations simultaneously and by dividing the overall large volume of data into parts processed in multiple parallel computations—that is, reducing the amount of calculation data for each parallel computation. This approach is also applicable to electrical-circuit calculations. The article presents a description of a method to accelerate electrical-circuit calculations through parallel computation of subcircuits demonstrated using the example of a four-phase step-up semiconductor converter with floating commutation.</p>