A more exact value for the optimal depth of the pipeline based on the recurrent formula, determining the processor performance for implementation of the applications involving branches and the data dependencies in execution trace as a function of depth (number of phases) of its pipeline, has been obtained.