паралельні обчислення

HIGH-PERFORMANCE COMPUTING METHOD FOR INVESTIGATING HEAT CONDUCTION PROCESSES IN COMPOSITE MATERIALS

This study explores an approach to modeling heat conduction in composite materials using high-performance computing methods. The solution of a complex mathematical problem is decomposed into multiple independent processes through the MPI method to maximize computational efficiency. The model structure accounts for the properties of the composite material and its components. The finite element method (FEM) is applied to study the thermal properties.

Врахування особливостей графічного процесора в процесі створення засобів автоматичного розпаралелення програм

Встановлено проблемні аспекти виконання паралельних алгоритмів на графічному прискорювачі. На основі встановлених особливостей запропоновано алгоритм генерації програм для графічного процесора.

This paper describes the main problems in implementing parallel algorithms on graphics accelerators. Based on the established features proposed algorithm generating programs for the GPU.

The parallel algorithm for solving problems of elasticity

Domain decomposition algorithm for solving problems of elasticity based on parallel computing is considered. The global system of equations for the entire domain is not formed and is represented by local matrices and vectors for subdomains using Boolean matrices of connectivity. The system of linear equations is solved by modified conjugate gradient method. The algorithm is implemented with C ++ using parallel MPI library. The results of testing proposed approach for modeling example are included.

Parallelization computations during simulation of mechanical components of mems-based diakoptical approach

The article reviews the main factors that determines the relevance of the development of special resolution of mechanical systems equations. We propose to use the diacoptical approach to design mechanical components of MEMS.

Особливості програмної реалізації розпаралелення процесу побудови дискретних динамічних моделей

Analysis of characteristics of software implementation of parallelization of process of constructing of discrete dynamical models was conducted in this paper. SIMD-architecture was used for the task of parallelization. Technology CUDA and GPU NVIDIA was used for this software implementation.