Code Optimisation and Performance Tuning for Xeon Processors (October 2016)

From SNIC Documentation
Revision as of 12:21, 29 September 2016 by Thor Wikfeldt (PDC) (talk | contribs) (Created page with "{{training event info |description=Advanced Programming techniques for performance improvement of numerical simulation codes |fields=Computational Science |event type=Course |loc...")
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to: navigation, search
Name Code Optimisation and Performance Tuning for Xeon Processors (October 2016)
Description Advanced Programming techniques for performance improvement of numerical simulation codes
Type of event Course
Location Stockholm (PDC)
Start date 2016-10-30
End date 2016-11-02

Course overview

Numerical algorithms have to be implemented very efficiently in order to meet performance requirements. As a consequence, knowledge of performance optimisation techniques is important for application developers, who, in the past, could trust compilers and other software tools to unleash the computational power of their systems. This training provides an introduction to performance optimisation techniques for numerical simulation programs.

Starting from current computer architectures, methods for performance tuning will be discussed. The efficient usage of memory hierarchies and arithmetic logic units are central to this and caches, loop optimisation and vectorization. This training will also cover further topics involving the optimisation of parallel programs, such as the use of OpenMP on shared memory computers.

Prerequisite knowledge

You will need to have an account on the PDC systems in order to attend and participate in the computer labs for the course. Please apply for an account in good time before the course if you do not already have a PDC account. The programming language that will be used in the course will be specified soon.

To attend the course, you need a background in programming in C, C++ or Fortran. In addition, knowledge of programming with OpenMP will be useful, although it is not compulsory for attendance.

Course web page and registration

To register for the course, see: [1].

Please note that the number of participants will be limited, so please register early to guarantee a place. Registration will close three weeks before the course.

What to bring

You need to bring a laptop to the course (for accessing PDC's computers).