Scientific IT Services supports ETH research
Launching and setting up a new division for ETH research.
Scientific IT is a valuable tool used in many research areas. Alongside actual research tasks such as developing new algorithms and computer models, IT services have increasingly become an important part of Scientific IT. This includes operation of servers for scientific computing as well as writing and maintaining software for research. Until recently, these kind of scientific IT services were mainly provided in institutes and groups. In order to bundle these services, the IT Services have established the new division Scientific IT Services (SIS).
High Performance Computing
High-performance computing was the first area in which such tasks were performed centrally. Already as far back as 2004, the IT Services started operating the first central Beowulf cluster „Hreidar“. The service was developed further in 2007 when ITS established the HPC group and in the next year the first generation of the ETH computer cluster „Brutus“ was put into operation. Currently the HPC group runs a third generation Brutus cluster. It is used by researchers from almost all departments of ETH Zurich. To meet the rising demand, ITS is planning to build a new compute cluster called „Euler“ which will complement Brutus.
Scientific software, data and workflows
HPC was the first area where researchers have expressed the desire for professional assistance with scientific IT services but
did not remain the only one. In the field of life sciences, data volume and complexity have increased exponentially due to new and more accurate methods of measurement, increased use of theoretical, computer-assisted methods and the systems biology approach. Accordingly, the task of analyzing, visualizing and publishing such a large amount of data became increasingly costly. At that point, the Center for Information Sciences and Databases (CISD) was established in 2006 to support researchers by writing scientific software and in particular with data and workflow management. In the same context, in 2009 under similar circumstances, SystemsX.ch initiated the SyBIT Project (Systems Biology IT) in order to organize scientific IT support for SystemsX research projects throughout Switzerland. Also at ETH, SyBIT staff members have since made significant contributions in building complex data analysis pipelines, e.g. in the field of mass spectrometry and „high content screening“.
Dr. Olivier Byrde, HPC, (left) holding a computer node; Dr. Bernd Rinn and Manuel Kohler with a picture showing the interior of a sequencing machine.
The new ITS division, Scientific IT Services, will bundle these services and make them available to a larger group of researchers, also in fields other than life sciences. The figure below shows how SIS is structured. The range of services will be further developed in collaboration with ITS scientific clients. SIS puts emphasis on the integration of services and will, wherever possible and appropriate, build on already existing services and collaborate with other service providers.
Access to the offerings of Scientific IT Services
SIS services are available to all ETH researchers on the basis of a cost contribution. This is implemented as a shareholder model similar to the one used by the HPC cluster Brutus, however shareholders pay for services by SIS professionals instead of computing power. The departments D-BSSE and D-BIOL decided already in the design phase to participate in SIS with their research groups in this form. Other departments and units have also expressed an interest in participating. In addition to the shareholder model SIS also provides services on a project basis.
During the last months the conceptual foundations of the Scientific IT Services division have been developed. Now the actual build-up is in full swing. If you would like to get further information on the services of SIS, or if you are interested in participating, please contact Dr. Bernd Rinn.
Dr. Bernd Rinn, Division Head SIS