The book "Scalable Data Management for Future Hardware" provides practical insights into nationwide research and the latest developments and shows how companies can make optimum use of modern hardware technologies, future storage technologies and high-speed networks to make database solutions more powerful.
Conventional database systems are reaching their limits
Data are the backbone of the digital economy: from Industry 4.0 and medical assistance systems to fraud detection software and digital humanities. Today, they are available in huge quantities and can be analyzed using machine learning methods. However, in order to successfully exploit the potential of artificial intelligence (AI), the data must also be managed sensibly.Rapid technological development is presenting data management with new challenges, and conventional database systems are reaching their limits when it comes to highly flexible queries, new types of data models or processing huge amounts of data in real time.
Research results from renowned database experts
The DFG Priority Program " Scalable Data Management for Future Hardware ", coordinated by TU Ilmenau, started right here: The aim was to bundle nationwide research activities in the field of database systems and related areas of operating systems and computer architectures in order to develop new system architectures that are better adapted to new application areas, make optimum use of the performance of modern hardware and meet the requirements of the digital future.The program was launched in 2016 by the Database Systems Section of the German Informatics Society (Gesellschaft für Informatik e.V.) together with Alfons Kemper (TU Munich, Chair of Database Systems), Thomas Neumann (TU Munich, Chair of Data Science and Engineering), Kai-Uwe Sattler (TU Ilmenau, Head of Databases and Information Systems Group) and Jens Teubner (TU Dortmund University, Chair of Database Systems) - internationally recognized experts in the fields of database systems, distributed systems and modern hardware architectures, whose research work has contributed significantly to the development of innovative data management technologies. From 2017 to 2024, scientists from Munich, Darmstadt, Berlin, Dortmund, Osnabrück, Magdeburg, Erlangen-Nuremberg and Marburg conducted research in two phases of this priority program. They were supported by an expert advisory board, which included Peter Boncz (CWI/VU Amsterdam), Franz Färber (SAP), Goetz Graefe (Google) and Theo Härder (TU Kaiserslautern), all’internationally renowned database experts.
Designing a future-proof IT infrastructure
Prof. Kai-Uwe Sattler, coordinator of the DFG priority program and Head of Databases and Information Systems at the TU Ilmenau:At a time when data-driven business models and artificial intelligence are becoming increasingly important, our research findings provide valuable insights for IT decision-makers, developers and researchers in the field of data management to meet the demands of practice and drive innovative ideas and challenging research questions.
In nine chapters, Sattler and his colleagues show how current and future hardware developments can optimize storage and data management solutions for big data and AI applications and how companies can adapt their IT infrastructure for the future using powerful database technologies. The topics covered include:
- Hardware-accelerated queries : Efficient use of co-processors such as GPU and FPGA to accelerate database operations.
- Memory and storage technologies in data management: Utilization of new technologies for non-volatile main memory and SSD arrays for data management
- Distributed data management with high-speed networks : Utilizing RDMA and programmable switches for faster data transfers.
Together, the editors and authors invite researchers in academia and industry to explore the latest developments in database management, to fully exploit the potential of modern data processing and thus achieve competitive advantages, reduce costs and future-proof their IT infrastructure.
The book "Scalable Data Management for Future Hardware" is freely available as an open access publication and enables broad use in science and industry: https://doi.org/10.1007/978-3-031-74097-8
