Parallelization of numerical methods

Course guarantor:

Jiří Fürst

Lecturer:

Jiří Fürst, Tomáš Oberhuber, Michael Valášek

Tutor:

Jiří Fürst, Tomáš Oberhuber, Michael Valášek

Supervisor:

Department of Computer Systems

Synopsis:

The content of the course is an explanation of numerical methods for solving mathematical models with a focus on their parallelization and the use of these methods in QC. Students are also introduced to modern trends in the field of solving these problems.

Requirements:

Syllabus of lectures:

1. Practical aspects of parallel programming on CPU and GPU.

2. CUDA language.

3. Hybrid CUDA-MPI programming, parallelization in Python.

4. Formulation of multibody system models.

5. Approaches to parallelization of multibody systems.

6. Methods for parallelization of multibody systems (including DEM, SPH, molecular dynamics. . . . ).

7. Formulation of models of elastic solids.

8. Discretization of models of elastic bodies.

9. Parallelization methods for models of elastic solids.

10. Formulation of fluid models.

11. Discretization of fluid models.

12. Methods of parallelization of fluid models.

13. QC simulators.

Syllabus of tutorials:

Not filled yet.

Study Objective:

The content of the course is an explanation of numerical methods for solving mathematical models with a focus on their parallelization and the use of these methods in QC. Students are also introduced to modern trends in the field of solving these problems.

Study materials:

1. Robey, R., Zamora, Z.: Parallel and High Performance Computing

ProQuest Ebook Central 2021

https://ebookcentral.proquest.com/lib/cvut/detail.action?docID=6683526

2. Scott, L. R., et al.: Scientific Parallel Computing

Princeton University Press, ProQuest Ebook Central 2005

https://ebookcentral.proquest.com/lib/cvut/detail.action?docID=7174828

3. Hwu, W. M. W., Kirk, D. B., Hajj, I. E.: Programming Massively Parallel Processors: A Hands-on Approach

Morgan Kaufmann 2022

ISBN 978-0-323-91231-0

4. Nelli, F.: Parallel and High Performance Programming with Python: Unlock parallel and concurrent programming in Python using multithreading

CUDA, Pytorch and Dask, AVA 2023

ISBN 10-9388590732

5. CUDA Programming Guide

https://docs.nvidia.com/cuda/cuda-c-programming-guide/index.html

Note:

Information about the course and teaching materials can be found at https://courses.fit.cvut.cz/QNI-PNM

Further information:

https://courses.fit.cvut.cz/QNI-PNM

No time-table has been prepared for this course

The course is a part of the following study plans:

• Quantum Informatics (compulsory elective course)