Answers are grounded in this paper and cite their sources.
This article provides a review of publications devoted to the features of using parallel computing systems for solving problems in continuum mechanics using the finite element method. These publications discuss the conditions for the successful application of algorithms and software developed for the efficient operation of parallel computing systems. Methods for accelerating the solution of a problem by taking into account its specifics are studied and estimates of their effectiveness are given. The issues of application are considered: iterative methods with preconditioning for solving sparse systems of linear algebraic equations with an irregular structure, both symmetric and asymmetric; finite element approximations on unstructured and quasi-structured computational grids; pipelineization of the computational process during spatial decomposition of the computational domain; a parallel computing method with an explicit and implicit time integration scheme for large-scale problems in the mechanics of structures and solids; finite element modeling of contact-impact problems using a graphics processor; matrix-free approach to finite element calculations intended for graphic processors; an explicit method of asynchronous step-by-step parallel calculations to reduce computation time when refining the mesh in a certain local area; discrete and finite element method and graphic processor for solid particle interaction modeling, including contact detection, force calculation and information updating; methods for parallelizing cyclic sections of sequential programs, based on determining the possibility of parallel execution of loop iterations, searching and parallelizing linear sections within loops with an unknown number of iterations.
No takes yet. Share an insight, caveat, or question.
Кибец et al. (2023) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: