Course - Supercomputing, Introduction - TMA4280
TMA4280 - Supercomputing, Introduction
About
Examination arrangement
Examination arrangement: Portfolio assessment
Grade: Letters
| Evaluation | Weighting | Duration | Grade deviation | Examination aids |
|---|---|---|---|---|
| Arbeider | 40/100 | |||
| Skriftlig eksamen | 60/100 | 4 timer |
Course content
The first part of the course gives an overview of modern computer architectures, as well as an introduction to the use of MPI for parallel computations. The rest of the course focuses on the choice of numerical algorithms as well as tuning and implementation of the numerical algorithms in order to achieve high performance both in a single-process and a multi-process context. Special attention is given to basic vector- and matrix-operations, direct and iterative solution of systems of equations, and numerical solution of partial differential equations.
Learning outcome
1. Knowledge. The student has a basic understanding of modern computer architecture and trends within the development of new computing systems. In particular, the student has an understanding of what can effect the performance of computational tasks (i.e., the number of floating point operations per second), and what can be done to optimize the performance. The student can identify how to decompose a global computational problem into sub-problems suitable for parallel processing, and can model the performance for selected algorithms. The student is familiar with the link between modern computer architecture and the development of efficient numerical algorithms for scientific and technical computation.
2. Skills. The student is able to implement numerical algorithms using a high level programming language, and with relevance to science and technology, both on a single processor and using multiple processors. The student can use functions from the communication library MPI to perform local and global communication tasks. Furthermore, the student can measure the performance of an algorithm and analyze the performance measurements with respect to scalability as the problem size and the number of processors vary. The student is familiar with different tools like automated partitioning tools and optimized routines for basic linear algebra operations (BLAS).
Learning methods and activities
Lectures and voluntary exercises. In addition, one or more compulsory problems will be given. Portfolio assessment is the basis for the grade awarded in the course. This portfolio comprises a written final examination (60%) and one or more compulsory problems (40%). The results for the constituent parts are to be given in %-points, while the grade for the whole portfolio (course grade) is given by the letter grading system. Retake of examination may be given as an oral examination. The lectures may be given in English. If the course is taught in English, the exam will be given only in English. Students are free to choose Norwegian or English for written assessments.
Compulsory assignments
- Exercises
Recommended previous knowledge
Familiarity with numerical methods and some experience with programming in Fortran or C.
Course materials
Will be announced at the start of the course.
Credit reductions
| Course code | Reduction | From | To |
|---|---|---|---|
| SIF5077 | 7.5 |
Version: 1
Credits:
7.5 SP
Study level: Second degree level
Term no.: 1
Teaching semester: SPRING 2017
Language of instruction: English, Norwegian
-
- Mathematics
- Technological subjects
Department with academic responsibility
Department of Mathematical Sciences
Examination
Examination arrangement: Portfolio assessment
- Term Status code Evaluation Weighting Examination aids Date Time Examination system Room *
- Spring ORD Arbeider 40/100
-
Room Building Number of candidates - Spring ORD Skriftlig eksamen 60/100 2017-06-01 09:00
-
Room Building Number of candidates - Summer KONT Arbeider 40/100
-
Room Building Number of candidates - Summer KONT Muntlig eksamen 60/100
-
Room Building Number of candidates
- * The location (room) for a written examination is published 3 days before examination date. If more than one room is listed, you will find your room at Studentweb.
For more information regarding registration for examination and examination procedures, see "Innsida - Exams"