Research and development at LRZ

Whether it’s data from modelling and simulation or images of satellites or measured values from sensor data in nature: The environmental sciences produce a wide variety of digital data and often require technical support for the processing of data. The LRZ is exploring useful transmission and processing methods, as well as harmonising data from various sources. For example, for Artificial Intelligence applications or for developing systems for disaster mitigation. When it comes to environmental computing, the V2C focuses on the visualisation of simulation results.

The hardware for supercomputing continues to become more heterogeneous in order to achieve better energy-efficiency. Software needs to be adapted to systems so that they can compute efficiently. The LRZ has installed the Bavarian Energy, Architecture and Software Testbed or BEAST, for conducting research into innovative IT: This is where the future of computing is being prepared, e.g. tools for controlling and monitoring supercomputers, and strategies for reducing the energy requirements of computing.

Artificial intelligence accelerates calculations and expands the set of methods in all scientific disciplines. But how reliable is AI? What data can be used to train systems? And how can the energy requirements of AI systems be reduced? The Big Data & Artificial Intelligence (BDAI) team explores AI applications, data strategies and innovative AI technologies to provide researchers with specialist support.

A picture is worth a thousand words: Immersive virtual reality applications also promote understanding in the sciences and fascinating images ultimately support communication strategies. Specialists at the Centre for Virtual Reality and Visualisation (V2C) at the LRZ are investigating how to present research from a technical perspective. In collaboration with researchers and providers, they develop tools and workloads for recording and processing image data.

Research data may contain more information than planned: Therefore, it should be reliably archived, as easy to search as possible and ultimately reusable and accessible. The professional management of research data is also becoming increasingly important for efficiency reasons. The LRZ handles digital information in line with the FAIR data principles and develops tools for managing data in the area of high-performance computing as well as platforms for publishing research outcomes, such as the LRZ Fair Data Portal.

Storing data
Searching for data

In addition to tape and hard disk archives, the LRZ Cloud also stores research data. There, however, they require special protection against hacking and manipulation. In collaboration with research groups, the LRZ explores and develops tools and technology for securely storing research data or for creating a Trusted Research Environment, which includes, for example, technically implementing regulations for access rights to sensitive data or managing stored data.

How can openly accessible science be implemented from a technical standpoint? How can we guarantee the security of information in decentralised systems and infrastructures? And how do computing centres ensure their service quality? Research questions and projects of general importance have arisen as a result of LRZ’s certifications. Together with the Digitalverbund Bayern and universities, the LRZ is researching strategies and measures to secure digital information, to increase the resilience of computing centres, and to also maintain IT services even when crises strike, keyword: business continuity.

Efficient science and information management is becoming increasingly important for our research and innovation projects. The Research and Information Management Team develops and tests new tools for use within and beyond the LRZ. The team focuses on customising science management methodologies, a comprehensive Research Information System, and work on an open web search within the OpenWebSearch.EU project and Open Search Initiative.