2026-10-14T02:38:18Z https://api.figshare.com/v2/oai

oai:figshare.com:article/34046432 2026-10-01T14:45:08Z category_7 category_15 category_45 category_734 category_106 portal_87 item_type_3 month_year_10_2026

Applying parallel computing in numerical methods: theoretical and practical foundations for education Perizat Gapbarova (25157048) Nursaule Karelkhan (25157051) Gulmira Bakirova (25157054) Medicine Neuroscience Sociology Biological Sciences not elsewhere classified Science Policy Educational methodology innovative technologies numerical methods parallel computing student satisfaction Arts & Humanities Arts Media Communication Humanities Media & Film Studies Media & Communications Media Education Social Sciences Behavioral Sciences Educational Psychology <p>This study explores a teaching methodology combining theoretical and practical components with modern parallel programming technologies. The training process utilized parallel programming languages and tools such as OpenMP, MPI, and CUDA, as well as the SimGrid simulator, which provided students with the opportunity to master real-world high-performance computing technologies. The research evaluates the method’s impact on students’ knowledge, skills, and perceptions of the learning process using tests, questionnaires, and statistical analysis. The study involved 118 students from four Kazakhstani universities. Participants averaged 20 years of age. The main problem of the study is that traditional approaches to teaching numerical methods do not meet the needs of the development of information and communication technologies in Kazakhstan, where specialists with parallel and high-performance computing skills are needed. Results showed a significant improvement in student performance, with theoretical exam scores increasing by 38.1% and practical scores by 45.9%. Statistical analysis confirmed these gains, validating the method’s effectiveness. Surveys further indicated enhanced student perceptions of program usefulness, satisfaction, and material comprehension. The methodology integrates modern technologies and tools, significantly improving theoretical and practical knowledge in parallel computing.</p> <p>This study demonstrates that integrating modern parallel programming technologies, including OpenMP, MPI, CUDA, and SimGrid, into the teaching of numerical methods significantly improves students’ theoretical and practical learning outcomes. The findings provide a practical framework for modernizing computer science education in Kazakhstan and other contexts where high-performance and parallel computing skills are increasingly required.</p> 2026-10-01T14:45:08Z Dataset Dataset 10.6084/m9.figshare.34046432.v1 https://figshare.com/articles/dataset/Applying_parallel_computing_in_numerical_methods_theoretical_and_practical_foundations_for_education/34046432 CC BY 4.0