В статье представлен подход к разработке информационно-аналитической системы, помогающей исследователю решать вычислительно сложные задачи математической физики на суперкомпьютерах. Система автоматически строит схему решения задачи по спецификации пользователя, введенной им в режиме диалога. Схема включает наиболее подходящие математические модели для решения задачи, численные методы, алгоритмы и параллельные архитектуры, ссылки на доступные фрагменты параллельного кода, которые пользователь может использовать при разработке собственного кода. Построение схемы осуществляется на основе онтологии проблемной области «Решение вычислительно сложных задач математической физики», онтологии заданной предметной области и экспертных правил, построенных с использованием технологии Semantic Web. The paper presents an approach to the development of an information-analytical system that helps a researcher to solve compute-intensive problems of mathematical physics on supercomputers. The system automatically builds a scheme for solving the problem according to the user's specification entered by him in the dialogue mode. The scheme includes the most suitable mathematical models for solving the problem, numerical methods, algorithms and parallel architectures, links to available fragments of parallel code that the user can use when developing their own code. The construction of the scheme is carried out on the basis of the ontology of the problem area "Solving compute-intensive problems of mathematical physics", the ontology of a given subject area and expert rules built using the Semantic Web technology.

Рассмотрены вопросы использования экзафлопсных вычислений для решения прикладных задач. На основе обзора работ в этой области выделены наиболее актуальные вопросы, связанные с экзафлопсными вычислениями. Особое внимание уделено особенностям программного обеспечения, алгоритмам и численным методам для экзафлопсных суперЭВМ. Приведены примеры разработки новых и адаптации существующих алгоритмов и численных методов для решения задач механики сплошной среды. Сделан анализ наиболее популярных приложений The article deals with applied issues which arise when exascale computing are used to solve applied problems. Based on the review of works in this area, the most pressing issues related to exascale calculations are highlighted. Particular attention is paid to software features, algorithms and numerical methods for exaflop supercomputers. The requirements for such programs and algorithms are formulated. Based on the review of existing approaches related to achieving high performance, the main fundamentally different and non-overlapping directions for improving the performance of calculations are highlighted. The question of the necessity for criteria of applicability for computational algorithms for exaflop supercomputers is raised. Currently, the only criterion which is used, demands the absence of a significant drop in efficiency in the transition from a petaflop calculation to a ten-petaflop calculation. In the absence of the possibility of such calculations, simulation modelling can be carried out. Examples of development for new and adaptation of existing algorithms and numerical methods for solving problems of continuum mechanics are given. The fundamental difference between algorithms specially designed for exascale machines and algorithms adapted for exaflops is shown. The analysis of publications has showed that in the field of solving problems of continuum mechanics, the approach not associated with the development of new, but rather with the adaptation of existing numerical methods and algorithms to the architecture of exaflop supercomputers prevails. The analysis of the most popular applications is made. The most relevant application of exaflop supercomputers in this area is computational fluid dynamics. This is because hydrodynamic applications are rich and diverse field. The number of publications indicates that the involvement of high-performance computing now is available and in demand