ECTS credits ECTS credits: 3
ECTS Hours Rules/Memories Student's work ECTS: 51 Hours of tutorials: 3 Expository Class: 9 Interactive Classroom: 12 Total: 75
Use languages Spanish, Galician
Type: Ordinary subject Master’s Degree RD 1393/2007 - 822/2021
Departments: Applied Mathematics, External department linked to the degrees
Areas: Applied Mathematics, Área externa M.U en Matemática Industrial
Center Faculty of Mathematics
Call: First Semester
Teaching: With teaching
Enrolment: Enrollable | 1st year (Yes)
Programming parallel computers. Parallelizing classical algorithms in matrix analysis and domain decomposition in discretized problems.
1. History and necessity of parallel computing.
2. Parallel architectures.
3. A first parallel program after a sequential program.
4. An application: numerical integration.
5. Collective communications.
6. Grouping data in communications.
7. Design of parallel algorithms.
8. Measuring the performance of parallel programs.
9. Parallelization of matrix-vector and matrix-matrix products.
10. Parallelization of classical methods for solving linear systems.
11. Parallelization of finite difference methods.
12. Domain decomposition methods for discretized problems.
13. Programming shared memory machines.
14. Combining MPI and OpenMP.
Parallel Programming in C with MPI and OpenMP. Michael J. Quinn (McGraw-Hill Science/Engineering/Math, 2003).
Introduction to Parallel Computing, Second Edition, by Ananth Grama, Anshul Gupta, George Karypis, and Vipin Kumar (Addison -Wesley, 2003).
Parallel Programming with MPI, by Peter Pacheco (Morgan Kauffman Publishers, 1997).
Parallel Programming, by Barry Wilkinson and Michael Allen (Prentice Hall, 1999).
CG1 Have knowledge that provide a basis or opportunity for originality in developing and / or applying ideas, often within a research context, knowing how to translate industrial needs in terms of R&D in the field of mathematics Industrial;
CG3 Being able to integrate knowledge in order to state opinions using information that even incomplete or limited, include reflecting on social and ethical responsibilities linked to the application of their knowledge;
CE4: Being able to select a set of numerical techniques, languages and tools, appropriate to solve a mathematical model.
CE5: Being able to validate and interpret the results, comparing them with visualizations, experimental measurements and functional requirements of the physical engineering system.
CS2: To adapt, modify and implement software tools for numerical simulation.
These skills are worked on exercises proposed in class, where we start from a sequential problem to transform it into a parallel one. In practicing exercises students will similarly work on their own on other problems of that type, and the assessment is carried out verifying these skills from the deliverables, taken into account with equal weight, that is, 0.2 for each of them.
Web portal. Guided exercises for the student to reinforce his/her knowledge of parallel programming.
Assessment of students will be done through assignments (1-3) and exercises that they will deliver, all of them equally weighted. Skills CG1, CG3 and CE4 are emphasized in exercises, and CE4, CE5 and CS2 in the assignments. Same thing second opportunity.
10 h (theory) + 20 h (practise)
Non-class hours: 45 h
Grand total: 75 h
Reserve time periodically for studying. Code all proposed examples from scratch. Use recommended books.
Jose Antonio Alvarez Dios
Coordinador/a- Department
- Applied Mathematics
- Area
- Applied Mathematics
- Phone
- 881813353
- joseantonio.alvarez.dios [at] usc.es
- Category
- Professor: University Lecturer
| Wednesday | |||
|---|---|---|---|
| 12:00-13:30 | Grupo /CLE_01 | Spanish | Computer room 5 |