Supercomputing at the leading edge
It is the mission of the Gauss Centre for Supercomputing (GCS) to offer computing power of the highest performance class for computational sciences and engineering at its three member sites in Garching (Leibniz Supercomputing Centre, LRZ), Juelich (Juelich Supercomputing Centre, JSC), and Stuttgart (High Performance Computing Center Stuttgart, HLRS). To ensure a most efficient utilisation of these highly valuable resources, GCS provides its users with worldleading support, education, and dissemination of best practices and methods in simulation science. GCS aims, in particular, at innovative and scientifically challenging large scaleprojects that cannot be carried out within smaller infrastructures. Such projects will also benefit most from the existing successful support structures within the GCS and from their continuous synchronisation and optimisation.
State-of-the-art systems
The GCS offers a highest-level computing and networking infrastructure. At JSC the first European PetaFlop Supercomputer, a 72 rack IBM Blue Gene/P system is available; HLRS provides the first installation step of HERMIT based on CRAY XE6 with AMD Interlagos and Gemini Interconnect with more than 1 PetaFlop/s; and LRZ offers its new IBM iDataPlex as a migration system for its upcoming Petascale system. The architectures of these machines are complementary and can accommodate a broad spectrum of applications. The systems within the GCS are continuously upgraded in a round robin fashion.
Large-Scale Projects
Large-scale projects and highly scalable parallel applications are characterised by large computing time requirements, not only for short time frames, but frequently for longer time periods. The GCS centres will guarantee that large-scale projects can fully utilise the assigned resources, if necessary at the expense of normal or smaller projects.
Projects are classified as “Large-Scale”, if they require more than 5% of the potentially available CPU cycles on a member centre’s high-end system, i.e.
- More than 70 Mill. core-hours at JSC (corresponding to 24 rack months IBM BG/P).
- More than 30 Mill. core-hours at HLRS (HERMIT Step 1).
- More than 4 Mill. core-hours at LRZ (IBM iDataPlex, Migration System for SuperMUC)
For these large-scale projects a competitive review and resource allocation process is established by the GCS. Requests for resources below these limits will be accepted by the individual member centres. Requests above these limits will be handed over to GCS, handled according to the joint procedures, and will be reviewed in the national context. Projects of smaller scale will be still handled by the local GCS member site.
A “Call for Large-Scale Projects” is published by the Gauss Centre twice a year. Dates for closure of calls are usually at the end of February and at the end of August of each year.
Eligible are applications from publicly funded academic and researchinstitutions, e.g., universities, Max-Planck Society and Helmholtz Association.
The deadline for the current call is 31 August 2011, 17:00.
The next call will be issued in January 2012.
Answering the Call
Leading, ground-breaking projects should deal with complex, demanding, innovative simula-tions that would not be possible without the GCS infrastructure, and which can benefit from the exceptional resources provided by GCS. Application for a large-scale project should be performed by filling in the electronic application form that can be accessed from the GCS web page “Computing Time”:
http://www.gauss-centre.eu/computing-time
Note that the regular application forms of the GCS member centres can be reached from there, and users are encouraged to choose a centre that provides the architecture and software most appropriate for their programs.
The proposals for large-scale projects will be first reviewed with respect to their technical feasibility. Afterwards, they will be peer-reviewed for a comparative scientific evaluation. On the basis of this evaluation by a GCS committee the projects will be approved for a period of one year and given their allocations.
Criteria for decision
Applications for compute resources are evaluated only according to their scientific excellence.
- The proposed scientific tasks must be scientifically challenging, and their treatment must be of substantial interest.
- Clear scientific goals and verifiable milestones on the way to reach these goals.
- The implementation must be technically feasible on the available computing system, and must be in proportion to the performance characteristics of these systems.
- The Principal Investigator must have a proven scientific record, and she/he must be able to successfully accomplish the proposed tasks. In particular, applicants must possess the necessary specialized know-how for the effective use of high-end computing systems. This has to be proven in the application for compute resources, e.g. by presenting work done on smaller computing systems, scaling studies etc.
- The specific features of the high-end computers should be optimally exploited by the program implemen-tations. This will be checked regularly during the course of the project.
Further Information
For further help please contact the member sites via
http://www.gauss-centre.eu/contacts
Current call as pdf:
http://www.gauss-centre.eu/datapool/page/71/GaussCall6.pdf
Comentarios