High Performance Computing

The goal of the High Performance Computing specializations is to provide theoretical and practical knowledge necessary for solving intricate problems in a wide area of natural and socioeconomic sciences using complex numerical models requiring large computational resources. Students will get familiar with theoretical aspects of the complex model design, numerical methods for their solution, and fundamental limitations of underlying computational systems. They will learn the architecture of distributed supercomputing systems, and software for modelling, implementation, optimization, testing and deployment of developed modes. Students will gain necessary knowledge to assemble high performance systems and predict their performance on real-world problems. They will understand the principles and intricacies of parallel and distributed programming. The will learn how to apply gained knowledge in solving extensive engineering problems. 

• The graduates have good insight into mathematical modelling of complex systems, numerical methods, parallel programming principles, and architectures of large computing facilities. They can design, implement, optimize and validate complex numerical models on workstations and superocmputing systems. They can integrate hardware and software components into robust units, optimize their configuration and predict performance and power consumption.  
• The graduates are ready for design work, operation and management of high performance computing systems, further for research and development work in the area of modelling, implementation, optimization and validation of complex problems from various fields including physics, biology, chemistry, artificial intelligence, big data or economics. They understand principles of project control and are able to apply them at the development, design and in application of computing systems. They are able to design and carry out an experiment, analyse and interpret the data, work individually or in a team, present the results in written or oral form and further educate themselves. Their adaptation is easy even into as diversified areas as demanding scientific and high performance computations and simulations.
• A carrier is possible in professions like system programmer, computer systems network designer and manager, system integrator, development specialist in design and rapid prototyping of computationally intensive models and applications. Graduates in this branch of study will use their education in development and research divisions of companies dealing with computer hardware and software and further in a huge number of companies, which needs specialists for high performance computing, state and local bodies, in the army, in education and health institutes, and practically in any industrial enterprise.

State Exam in Information Technology and Artificial Intelligence, specialization Supercomputing consists of the following parts:

• presentation and defense of master's thesis,
• oral exam, which combines the basic themes contained in the courses profiling the basis of Information Technology and Artificial Intelligence (Theoretical Computer Science, Statistics and Probability, Computer Systems Architectures, Artificial Intelligence and Machine Learning, Data Storage and Preparation, Functional and Logic Programming, Parallel and Distributed Algorithms, Modern Trends in Informatics),
• oral exam, which combines the basic topics covered in the courses profiling base in Supercomputing (Practical Parallel Programming, Parallel GPU Computations, Highly Computational HW/SW Codesign, Matrix and Tensor Calculus, Data Transfer, Computer Networks and Protocols).

All parts of the state examination are held on the same date before the State Examination Board. The state exam can be taken by a student who has obtained the required number of credits in the prescribed composition necessary for the successful completion of the master's degree and has submitted the master's thesis in due time. The organization and course of the state examination are given by the corresponding internal standard of the faculty and by the relevant instructions of the program guarantor for state examinations.

• Analysis of Operational Data and Detection od Anomalies during Supercomputer Job Execution
• Simulation of Fracture Tests in Civil Engineering
• Acceleration of Axisymetric Ultrasound Simulations
• Development and Programming of Low Power Cluster
• Performance Analysis of IBM POWER8 Processors
• Interactive Cloth Simulation Accelerated by GPU
• Dynamic Load-Balancing in Parallel Applications
• Optimization of the Distributed I/O Subsystem of the k-Wave Project