Parallelization, Optimization and Debugging in High Performance Computing 

Faculty Faculty of Mathematics, Computer Science and Physics
Duration / ECTS-Credits 3 days / equal to 1 ECTS-Credit
Location University of Innsbruck
Period expected January / February 2020
Qualification Certificate from the University of Innsbruck
Cost* € 300.-
Further Information

Folder

Website

Administration
Division for continuing education
Innrain 52f (Bruno-Sander-Haus)
6020 Innsbruck
Tel.: +43 512 507–39402
  weiterbildung@uibk.ac.at
Enrolment currently closed - go here to the list of prospects (german)


Course Director

Prof. Thomas FahringerUniv.-Prof. DI Dr. Thomas Fahringer
Department of Computer Science

 

 

 

 

 

Philipp GschwandtnerPhilipp Gschwandtner, MSc PhD
Department of Computer Science


 


  

 

Content

High Performance Programming with MPI

The participants learn to design, implement and optimize parallel programs using the MPI programming model for their own use cases in academia and industry. They study several approaches to solving a problem with MPI and selecting the best option with regard to performance, communication overhead, readability, and maintainability.

This course will be taught in English. 

 

Target Group

The target audience includes pre- and postdoc researchers at the University of Innsbruck, who require high computational power or performance-oriented programming for their scientific problems; furthermore, employees working at companies with high computational needs or complex problems that require parallelization; in addition, depending on registration numbers, master students of relevant curricula (e.g. Computer Science, Mathematics).

 

Qualification profile

Graduates are qualified to devise parallel programs for their scientific problems in academia and industry and produce high-quality implementations. They are able to evaluate several programming model and hardware architecture candidates, consider the respective advantages and disadvantages of each technology with regard to their specific use case, and choose the best option for realization.

Moreover, they are qualified to optimize their parallel programs analytically and empirically with regard to the specific properties of both their scientific problem and the chosen technologies in order to ensure high performance and efficiency.

 

Requirements

Basic programming skills in C, C++ or Fortran are required, as well as working knowledge of Linux/Unix commands.

 

Contact and Information

Department of Computer Science
Philipp Gschwandtner, MSc PhD
Technikerstr. 21a
6020 Innsbruck
Tel.: +43 512 507-53233
  Philipp.Gschwandtner@uibk.ac.at

*Payment terms and cancellation conditions

Nach oben scrollen