Applying for Computing Time on HLRS Resources

Eligible scientists must be employed at universities or research facilities in Germany. Here, the nationality of the applicant (principal investigator, PI) does not play a role. Applicants from Europe but outside of Germany are encouraged to apply for time through PRACE. The project PI must have a proven scientific record (preferably a PhD or comparable degree) and must be able to successfully accomplish the proposed tasks. For regular projects (those requiring less than 35 million core hours per year), there is a rolling call, with applications accepted at any time.

HLRS will replace Hazel Hen with Hawk in autumn 2019, and the new machine is scheduled to be available in October, 2019. Hawk provides 5,000 nodes, each one equipped with the next generation AMD Rome processors, offering 128 cores per node. 4,800 nodes have 128 GByte of main memory, 200 nodes provide 256 GByte. The majority of current requested allocations will be used on Hawk. Applicants will need to port their applications to the new system during summer, 2019. HLRS support is available and required support should be outlined in the project application.

During the first part of the allocation, HLRS offers its Cray system Hazel Hen, a XC40 based on Intel Haswell Processors and the Aries interconnect, with 7.4 PF/s peak performance. The 7,712 compute nodes provide 128 GByte main memory each and a total of 185,088 cores. As Hazel Hen will be reduced in size to provide space for Hawk, only a quite limited part of the allocation will be possible on Hazel Hen.

Further reporting obligations

Each project has to submit an Annual Report, which has to be provided in the first quarter of the following year. At the same time a publication list containing all publications based on research results achieved through the compute project has to be submitted per Email.

How to acknowledge granted GCS resources

GCS large-scale projects

The authors gratefully acknowledge the Gauss Centre for Supercomputing e.V. (www.gauss-centre.eu) for funding this project by providing computing time on the GCS Supercomputer HAZEL HEN at Höchstleistungsrechenzentrum Stuttgart (www.hlrs.de).

Regular GCS projects

Please make sure to mention the use of the GCS infrastructure at HLRS by an acknowledgement in your publications e.g. like "The simulations were performed on the national supercomputer Cray XC40 at the High Performance Computing Center Stuttgart (HLRS) under the grant number <Acronym>/<Account ID>.".

Email address for further questions

gcs-project-coordination@hlrs.de