The scope of work of the Supercomputing for Industry team is in the development, implementation and optimization of the libraries for parallel computations and their application. Our researchers use in-house, open-source and commercial tools to solve real engineering problems such as flow modeling, structural mechanics and biomechanics.
The development of efficient parallel algorithms is initiated mainly by the industrial sector. This development utilizes results of basic research,of the Department of Applied Mathematics of the VSB-Technical University of Ostrava. This department performs highly successful research in the area of development of numerical and parallel scalable algorithms and their application to engineering problems in cooperation with the Department of Mechanics and the Department of Mechanics of Materials of VSB-TU Ostrava. Our researchers have also long-term collaborations with renowned local and international research groups and industrial partners (see Projects).
In the implementation phase an analysis of efficiency of the implementation, focused on identifying the “bottlenecks” and its optimization for a specific architecture are performed. Programming languages and methods suitable for programming on massively parallel architectures (see High Performance Computing) together with existing optimized libraries (e.g. PETSc, Trilinos, MUMPS,LibMesh, Paraview, etc.) are used.
Our libraries are applied together with other open-source and commercial software for the solution of engineering problems from areas such as Computational Fluid Dynamics, Computational Structural Dynamics and biomechanics. In this research we apply our knowledge, experience and skills of already finished projects (see Portfolio).
We also provide customized solutions for our clients to fully utilize their computer architecture. Main objective of this service is to provide time and cost effective solution complementary to experimental measurements (see Service).