Introduction to Performance and Correctness Analysis for Parallel Programs

Welcome to the HPC homepage providing course and additional online information.
The Campus System gives an up-to-date account of lecture times and room assignments for the lecture and the excercises.

---PLEASE SIGN UP FOR THE LECTURE AND EXERCISES IN CAMPUS-OFFICE---

Current schedule: lecture starts on 17. of October 2013

  • Tuesday lectures take place in room (2356|055) from 14:15 to 15:45.
  • Thursday lectures take place in room (2356|055) from 10:15 to 11:45.
  • Exercises will be held during the lecture time (thursday) with lecture and excercise alternating every two weeks.
  • The lecture can be found in Campus-Office here

Description

Increasingly-complex scientific and industrial simulations need an increasing degree of parallelism to exploit the power of high-performance computers. This increase in parallelism introduces new complexity to application programming. The educational objective is to enable the student to detect and get an understanding for performance issues and programming mistakes within parallel applications. The
student is introduced to common tools for (semi-) automatic analysis.

Beschreibung

Um immer komplexere wissenschaftliche Simulationen effizient auf heutigen und zukünftigen Großrechner laufen zu lassen, ist ein steigender Grad an Parallelität nötig. Der steigende Grad an Parallelität geht mit steigender Komplexität der Anwendungsprogrammierung einher. Lernziele sind das Erkennen und Verstehen von Performance-Problemen und Quellen von Programmierfehlern in parallelen Anwendungen. Zur (semi)- automatischen Analyse werden verschiedene Werkzeuge vorgestellt.

News and materials:

Please use L²P: Learning and teaching portal of RWTH Aachen University to get news and materials for the lecture and exercises.

Slides for 2 minute madness can be found here.

  • No labels