Computing method for solving large scale matrix problems

In this paper it is shown a possibility of creating efficient computing technology for solving matrix-based scientific, engineering, economic, and other problems. It is based on using the new class of hypercomplex systems opened by W.F. Hamilton but formed by a procedure which is different from the,procedure proposed by his followers. This method allows, for example, to reduce multiplication and summation of two large size matrices to multiplication and summation of only two whole numbers and, in general, to use a single real number as an operand instead of matrix. Transforming matrices and vectors into whole numbers can be performed on the basis of parallel multiprocessor system.