News Release

Blue gene active storage boosts I/O performance at JSC

Supercomputing Conference 2013 (17-22 Nov.) in Denver: Scientists from Forschungszentrum Juelich announce the successful installation of a first Blue Gene Active Storage system worldwide

Peer-Reviewed Publication

Forschungszentrum Juelich

Supercomputer JUQUEEN

image: The BGAS system is attached to the supercomputer JUQUEEN installed at Jülich Supercomputing Centre (JSC). This Blue Gene/Q system is currently the most powerful system in Europe. view more 

Credit: Forschungszentrum Jülich

Jülich / Denver, 18 November 2013 – The ever growing complexity of various simulation tasks not only require a continuous increase of computing power but also the capability of managing large amounts of data. Realisation of an active storage architecture and integration of non-volatile memory into Blue Gene/Q enables data intensive applications to exploit the performance of this highly scalable high-performance computing system by IBM. The BGAS system is the result of a close collaboration between Forschungszentrum Jülich and IBM in the framework of the Exascale Innovation Centre. It is attached to the supercomputer JUQUEEN installed at Jülich Supercomputing Centre (JSC). This Blue Gene/Q system is currently the most powerful system in Europe.

Active storage is an architectural concept that addresses the increasing costs of data transport between compute and storage systems. Therefore, computing power and storage are much more tightly integrated. In addition, in BGAS this is combined with the power of non-volatile memory technologies which feature not only high bandwidth but in particular very high access rates. The BGAS system at JSC consists of 32 nodes each connected to a newly designed PCIe card comprising 2 TBytes of SLC NAND flash memory. The BGAS nodes are interconnected within a 3-dimensional torus network. Furthermore, each node is connected to a Blue Gene/Q compute system and to an external storage facility. This architecture is particularly suitable for data intensive tasks and thus enables new research.

"Highly complex simulations such as in brain research generate large volumes of data that can no longer be processed by conventional methods. New I/O concepts are therefore one of the most important steps to achieve new research results using high-performance computers in these areas," says the director of JSC, Prof. Thomas Lippert.

New opportunities for neuroscience simulations

The opportunities of the newly installed BGAS had been immediately recognised by the computational neuroscientists at Forschungszentrum Jülich. "During extensive simulations of neuronal networks large amounts of data must be stored externally in order to be analyzed later," says Prof. Markus Diesmann. The director at the Institute of Neuroscience and Medicine is one of the developers of the freely available software NEST, which is used by neuroscientists worldwide. However, no typical HPC system does provide sufficient I/O capabilities to cope with the generated data that researchers would like to analyse. "BGAS can sustain the required data rates and provides the compute power to analyse this data," says Markus Diesmann. The new storage system additionally facilitates interactive access to the data.

A team led by the Jülich neuroscientist recently succeeded in simulating the largest neuronal network to date consisting of 1.73 billion nerve cells on the Japanese K computer (see press release of 2 August 2013). The result demonstrates that the neuroscientists can make full use of existing petascale supercomputers.

###

Further information:

Press release of 2 August 2013, "Largest neuronal network simulation to date achieved using Japanese supercomputer": http://www.fz-juelich.de/SharedDocs/Pressemitteilungen/UK/EN/2013/13-08-02LargestSimulation.html

Blue Gene Active Storage

Workshop "Blue Gene Active Storage", January 2013: http://www.fz-juelich.de/ias/jsc/EN/Expertise/Services/Documentation/presentations/presentation-bgas_table.html

Exascale Innovation Center (EIC)

The Exascale Innovation Center (EIC) is a collaboration of JSC and IBM Böblingen on the development of hardware and software technologies for exascale computing with the goal of enabling operation of an exascale system around 2020. http://www.fz-juelich.de/ias/jsc/eic

Supercomputer JUQUEEN

The installation of JUQUEEN at JSC started in 2012. The system has been extended several times and reached a final configuration of 28 racks in 2013. The 458,752 processor cores provide a peak performance of 5.9 Petaflops. The system is purely used for academic research and is available to scientists coming from various research areas, including neuroscience, biology, energy and climate research as well as quantum physics. http://www.fz-juelich.de/ias/jsc/juqueen

Simulation software NEST

http://www.nest-initiative.org

Contact:

Prof. Dirk Pleiter
Jülich Supercomputing Centre (JSC)
Tel. +49 2461 61-9327
d.pleiter@fz-juelich.de

Press contact:

Tobias Schlößer
Tel: +49 2461 61-4771
t.schloesser@fz-juelich.de

Forschungszentrum Jülich…

… pursues cutting-edge interdisciplinary research addressing the pressing issues of the present. With its competence in materials science and simulation, and its expertise in physics, nanotechnology and information technology, as well as in the biosciences and brain research, Jülich is developing the basis for the key technologies of tomorrow. In this way, Forschungszentrum Jülich helps to solve the grand challenges facing society in the fields of energy and the environment, health, and information technology. Forschungszentrum Jülich is also exploring new avenues in strategic partnerships with universities, research institutions and industry in Germany and abroad. With more than 5,000 employees, Jülich – a member of the Helmholtz Association – is one of the large interdisciplinary research centres in Europe.


Disclaimer: AAAS and EurekAlert! are not responsible for the accuracy of news releases posted to EurekAlert! by contributing institutions or for the use of any information through the EurekAlert system.