Cray XC40, Parallel I/O and Optimization Courses

Enterprises & SME Research & Science
Cray XC40, Parallel I/O and Optimization Courses


Series of three Courses:

2015, Tuesday Oct. 20, 9:30 -17:30:

2015, Wednesday Oct. 219:00 - 17:30:
Efficient Parallel I/O

2015, Thursday-Friday, Oct. 22-239:00 - 17:30:
Optimization on Cray XC40

Please book the appropriate days in the registration form!


In August 2015, the Cray XC40 supercomputer Hornet at HLRS was upgraded to a new system named “Hazelhen" featuring 7724 compute nodes, each equipped with two 12 core Intel Haswell processors running at 2.5 GHz. Each node is equipped with 128 GB DDR4 memory and connected to the other nodes through the Cray Aries network. The peak performance amounts to 7.4 PFlops.

In order to help users running efficiently on this new large system, HLRS and Cray offers a workshops in October consisting of 3 parts.


The first day gives an overview on this new XC40 system. Specialists from Cray will talk about the hardware, best practices, and the new software enhancements, which will allow you to improve your performance on this new machine.
Target audience: If you are interested in updates or CRAY XC40 systems in general, you are welcome to attend this day.
Prerequisites: none

Efficient Parallel I/O:

The second day is dedicated to user parallel IO at scale, which becomes more and more important. High performance computing produces drastically increasing amount of data, at decreasing time scales. This workshop will cover parallel file systems, efficient data handling, and efficient parallel IO methods.
Target audience: Everyone who is welcome, who is interested in efficient reading and writing data on parallel systems.
Prerequisites: none

Optimization on Cray XC40:

The third and fourth day is mainly intended as a "Bring Your Own Code" (BYOC) workshop, and will cover the Cray Programming Environment, scientific libraries and profiling tools. Specialists from Cray will assist profiling and optimizing your programs.
Target audience are users running applications which are only scaling to a lower number of cores then expected or needed.
Prerequisites: In order to attend the BYOC workshop, you will have to bring your own code, and you will be required to do some work on your application before the workshop (details will follow). For these last two days, it is required that your have already an account on HAZELHEN and your application is already running on the system.

Information on access to Hazelhen from Germany and Europe is available through this link.

General Information


2015, Tuesday Oct. 20
9:00 - 9:30: Local registration,
9:30 -17:30: Cray XC40/HORNET at HLRS

2015, Wednesday Oct. 21
9:00 - 17:30: Efficient Parallel I/O

2015, Thursday-Friday, Oct. 22-23
9:00 - 17:30 (Thu.), 9:00 - 16:30 (Fri.): Optimization on Cray XC40




Stefan Andersson, Aniello Esposito and Michael Neff (Cray)


Online form not yet available


for registration is Sep. 20, 2015


Members of German universities and public research institutes: none.
Members of universities and public research institutes within Europe or PRACE: none.
Members of other universities and public research institutes: 120 EUR. 
Others: 400 EUR.
(includes food and drink at coffee breaks, will be collected on the first day of the course, cash only)

For days 2-4 programming experience in C or Fortran, some knowledge about parallel programming.


HLRS is part of the Gauss Centre for Supercomputing (GCS), which is one of the six PRACE Advanced Training Centres (PATCs) that started in Feb. 2012.
HLRS is also member of the Baden-Württemberg initiative bwHPC-C5.
This course is also provided within the framework of the bwHPC-C5 user Support. This course is not part of the PATC curriculum and is not sponsored by the PATC program.

Local Organizer

Rolf Rabenseifner
phone 0711 685 65530

Joerg Hertzer
phone 0711 685 65932

Cancelation Policy

If you cannot come to the course, please send an email to the organizer as soon as possible. This would allow us to accept additional participants from the waiting-list. There is no cancelation fee.
NO-SHOW: Registered persons that do not cancel and do not show up without any reasons are blocked for the next year on any of our workshops (because it is too expensive to produce unused copies of the slides for them).


Each participant will get a paper copy of all slides.
The MPI-1 part of the course is based on the MPI course developed by the EPCC Training and Education Centre, Edinburgh Parallel Computing Centre.
If you want, you may also buy copies of the standards MPI-3.0 (Hardcover, 17 Euro) and OpenMP (about 12 Euro).
An older version of this course with most of the material (including the audio information) can also be viewed in the ONLINE Parallel Programming Workshop.

Further courses

Shortcut-URLs of this course: