Basic compiler algorithms for parallel programmes

Parallel programs have become more and more important in recent years because of the rapid growth of information arrays and as a consequence the necessity of fast processing of large amounts of data. At the same time, traditional compilation methods developed for sequential programs do not guarantee correctness (sequential consistency) of compiler's transformations when applied to parallel programs. In this connection the article is devoted to the analysis of basic compiler algorithms for such programs allowing to apply parallelisation and optimisation methods for code creation.