Application for Projects on JUQUEEN
Regular calls and calls for large-scale-projects
The GCS/NIC-call is open for project leaders, whose affiliation is in Germany (or is a foreign office of a German institution). Also eligible are German scientists if they are working in an international organisation with significant German participation (e.g. CERN, ESA, ESO) and do not have a permanent position there. Interested researchers from abroad (within Europe) are invited to apply via PRACE.
Project applications for JUQUEEN may be submitted by any scientist qualified in his or her respective field of research. Computing resources are allocated on the basis of independent referees' reports. Apart from the scientific relevance of the project, an important criterion for the allocation of computing resources is that the project can make efficient use of the computer and use a large number of processors in parallel for the simulations.
Please consider the following notes on the use of JUQUEEN:
The resources of this leadership-class system will be provided for a small number of grand challenge projects that will be selected in accordance with strict scientific standards. Resources will be allocated to the projects within the framework of an international reviewing procedure. The following criteria must be met by the projects in order to be eligible:
- Scientific excellence.
- Clear scientific goals and verifiable milestones on the way to reach these goals.
- Preliminary studies that demonstrate the scalability of the program to very high processor numbers (at least 8192 cores) and proof of very good scalability at least up to 4096 processor cores (contact firstname.lastname@example.org).
- A detailed and clearly arranged work schedule in form of a table or Gantt chart.
- Well-founded and detailed demonstration of the required runtime of the program and the total required CPU time.
Please consult also the detailed technical guidelines for projects applying for JUQUEEN
The smallest mount of computing time that should be requested is 5 Mill. core hours.
Large-scale projects (35 million core hours or more) are peer-reviewed by a committee of the GCS. If approved they will get increased user support and preferential processing of their jobs.
Regular projects are peer-reviewed by a committee of the John von Neumann Institute for Computing (NIC) on behalf of GCS.
Computing time periods are yearly - with the possibility of application twice per year - and will begin on 1 May and on 1 November each year. The next computing time period will begin on 1 May 2015 and will end on 30 April 2016.
The next call for proposals will be announced at the end of January 2015. The deadline for the next application is 27 February 2015, 17:00.
Please, first prepare a detailed project description that you will have to upload to the electronic questionnaire as a pdf file. Follow the guidelines about form, content and size of the project description. Take particular care that
- the estimate of the applied computing time is verifiable by the reviewer,
- a plot is supplied that shows the scalability of the program under realistic production conditions, and that
- you provide information on computing projects and applications for computing time you have with other supercomputing centres.
If you wish to add supplemental material to the project description (unpublished manuscripts etc.), please collect that in a second pdf-file. Also this file you will have to upload to the electronic questionnaire.
Please fill in the electronic questionnnaire completely and attach the project description and - if necessary - the supplemental material as pdf files. Please also note that you have to provide technical information on your jobs, so have these data available (see).
Electronic questionnaire (open 30 January - 27 February 2015, 17:00)
Please make sure that your data are complete and check them. Finally, please print the application form and the PI agreement and send signed paper copies to the NIC coordination office at JSC.
NIC coordination office
Jülich Supercomputing Centre