On the stochastical and quasistochastical algorithms parallelisation in modeling problems

As it is known, the most of the mathematical modelling problems are connected with linear algebraic systems solving. The stochastical and quasistochastical method for L. A. S. solution are analysed in this article.
As it turned out this methods are asymptotical optimal in the class of the iteration methods as order of the systems tends to infinity. Besides some modifications of the offered methods possessed properties of unlimited parallelism.