EN
After a description of the basic concepts, a model of a parallel computer is formulated. The main results about complexity of parallel algorithms for the calculation of linear recurrence relations are described. Constant speed-up for the nonlinear case is also mentioned. The parallel algorithms for a system of linear equations with a tridiagonal, triangular and full matrix are discussed. The method of problem- and algorithm-decomposition is demonstrated. Asynchronous algorithms for an MIMD-type computer are defined, and with some numerical examples for linear and nonlinear equations, these ideas are demonstrated.