Applying the list method to the transformation of parallel algorithms into account temporal characteristics of operations

A program that is working inefficiently leads to inevitable losses in computer performance. These losses should be avoided or at least minimized. In order to do this we need to apply the approved research and development techniques and also equivalent algorithm conversions. In the present paper we suggest a technique of parallel algorithm modification with the aim of improving its efficiency due to balanced load of processors. The technique itself consists in rearranging processes among processors and enlargement of algorithm operations. All information dependencies of an algorithm are preserved and algorithm performance time and the number of processors involved can only be reduced.