| module number | 200003 |
| examination number | 220424 |
| department | Department of Computer Science and Automation |
| ID of group |
2252 (Data-intensive Systems and Visualization)
|
| module leader | Prof. Dr. Patrick Mäder |
| term | summer term only |
| language | Deutsch |
| credit points | 5 |
| on-campus program (h) | 45 |
| self-study (h) | 105 |
| obligation | elective module |
| exam | examination performance with multiple performances |
| details of the certificate | Das Modul Parallel Computing mit der Prüfungsnummer 220424 schließt mit folgenden Leistungen ab:
- akternatvie semesterbegleitende Prüfungsleistung mit einer Wichtung von 60% (Prüfungsnummer: 2200630)
- alternative semesterbegleitende Prüfungsleistung mit einer Wichtung von 40% (Prüfungsnummer: 2200631)
Details zum Abschluss Teilleistung 1:
- or multiple written tests consisting of multiple-choice and free-form questions evaluating the professional competence in the course's topics
- preferably conducted digitally via Moodle and on the student's device
- final results may be scaled or individual questions may be excluded depending on best performing percentile of students
- students must register via thoska for this exam, typically within the 3rd and 4th week of the semester
Details zum Abschluss Teilleistung 2:
- one or multiple assignments to be solved individually at home and turned-in via Moodle at a defined due date announced with the task
- assignments are may be accompanioned by either a short physical, oral presentation and discusion in front of the peer group OR a short video presentation; students will be informed about the selected form upon announcing assignment topics
- result determined as average across the evaluated solutions to the assignments
- students must register via thoska for this exam, typically within the 3rd and 4th week of the semster
|
| link to Moodle course |
https://moodle2.tu-ilmenau.de/course/view.php?idnumber=200003"
|
| teacher | Prof. Dr. Patrick Mäder |
| signup details for alternative examinations | Dieses Modul enthält mindestens eine alternative semesterbegleitende Abschlussleistung. Bitte beachten Sie, dass diese in der Regel schon zu Beginn des Semesters, in dem diese angeboten wird, angemeldet werden muss.
Über die Details und Zeiträume dazu werden Sie vom Lehrenden und/oder dem Prüfungsamt informiert. Fragen Sie gegebenenfalls unbedingt beim Lehrenden nach.
This module contains at least one alternative exam part. Please note that this must usually be registered at the beginning of the semester in which it is offered.
The lecturer and/or the examination office will inform you about the details and time periods. If necessary, be sure to ask the lecturer. |
| maximum number of participants | |
| previous knowledge and experience | - basic programming skills in C are beneficial
|
| learning outcome | Professional Competence mostly gained in lectures and evaluated in the oral exam:
- Students have knowledge about the fundamental concepts and terminology of parallel systems.
- Students have knowledge about different taxonomies to classify parallel hardware and the advantages and disadvantages per class.
- Students know different methodologies for decompisting, agglomerating, and mapping a given problem into a set of parallel executable tasks.
- Students know and can apply different synchronizatio techniques for parallel programms.
- Students have knowledge about different metrics for evaluating parallization success and are informed about best practices and problems when profiling parallel software.
Methodological Competence mostly gained in seminars and evaluated in the aPl (assignments):- Students gained the ability to implement parallel programms on different hardware plattforms including the ability to analyze and decompose a given problem for parallel computing.
- Students are able to independly develop individual parallel implementations to a given problem and are able to judge and compare the quality and succes in terms of parallization.
- Students gained the ability to evaluate and troubleshoot parallel programs.
- Students gained the ability to use development tools and computational resources (e.g., cloud computing instaces) for programming parallel programms.
Social Competence gained through lectures and seminars:- Students can discuss advantages and disadvantages of different deep learning approaches among each other and with their lectureres.
|
| content | The goal of this master-level course is giving a structured introduction into the concepts of parallel programming. Students will learn fundamental concepts of parallelization and will be able to judge the correctness, performance and construction of parallel programs using different parallelization paradigms (e.g. task parallelization, data parallelization) and mechanisms (e.g. threads, task, locks, communication channels). The course also provides an introduction to the concepts of programming and practical aspects of programming massively parallel systems and cloud computing applications (using Amazon AWS). At the end of the course, students shall be able to design and implement working parallel programs, using shared memory programming on CPU (using pThreads and OpenMP) and GPU (using Cuda) as well as distributed memory programming (using MPI) models. The concepts conveyed in lectures are deepened by practical programming exercises.
The following topics will be covered through lecture and seminar:
- Fundamentals of parallel algorithms
- Decomposition, Communication, Agglomeration, and Mapping of parallel tasks
- Styles of parallel programs
- Shared-memory programming
- Processes, threads, and synchronisation
- pThreads
- OpenMP
- Hardware architecture for parallel computing
- Shared and distributed memory
- Flynn's Taxonomy
- Cache
Coherence - Interconnection networks und routing
- Distributed-memory programming
- Message passing programming
- MPI
- Analytical program models
- Amdahl's law, etc.
- Metrics
- Profiling
- Parallel algorithms
- Programming
massivly parallel systems
- GPU
und CUDA Programmierung - OpenCL
- Warehouse-scale computing
|
| media of instruction and technical requirements for education and examination in case of online participation |
- Lecture and seminar slide decks
- Assignments managed via Moodle
- Additional material, e.g., papers, source code excerpts, etc., and development tools
- All material will be shared via Moodle, accesible [HERE]
Technical Requirements
- personal computer required for all seminars and assignments
- ... with access to moodle.tu-ilmenau.de
- ... with access to colab.google.com
|
| literature / references | - Introduction
to Parallel Computing: Zbigniew J. Czech, Cambridge University Press (2017)
- Introduction
to Parallel Computing (Second Edition): Ananth Grama, Anshul Gupta, George
Karypis, Vipin Kumar, Addison Wesley (2003), ISBN 0-201-64865-2
- Programming
Massively Parallel Processors: A Hands-on Approach, D.B. Kirk and W.W. Hwu,
Morgan Kaufmann, 2. Ed. (2012)
- Parallelism
in Matrix Computations, E. Gallopoulos, B. Philippe, A.H. Sameh, Springer
(2015)
- Parallel
Programming, T. Rauber and G. Rünger, Springer (2013)
|
| evaluation of teaching | |
