Distribution of computations in hybrid computing systems when translating NORMA language programs

The methods of computational load distribution when translating programs from the nonprocedural (declarative) NORMA language into executable programs for various parallel architectures are discussed. Some brief characteristics of the NORMA language and the main features of the compiler for programs in NORMA language are given. The methods of automatic distribution of computational load when generating executable programs of the following types are described: OpenMP, NVIDIA CUDA, MPI+OpenMP, and MPI+OpenMP+NVIDIA CUDA. The problem of dynamic computational load balancing arising in the case of the heterogeneous computing environment MPI+OpenMP+NVIDIA CUDA is considered and a method of solving it is proposed. The results of practical application of the compiler for the programs in NORMA language for solving two different mathematical problems are given and the performance of the resulting executable programs is estimated for various parallel architectures.