Data Engineering & Cloud Systems
The Technical University of Munich (TUM) is an internationally renowned center of excellence for database engineering and cloud systems. Its reputation is evidenced by numerous prestigious scientific awards and collaborations with leading industrial and academic partners. The group's strength lies in its systems-centered approach, which has led to the development of widely-used data management systems such as the HyPer and Umbra. HyPer was successfully commercialized and later acquired by Tableau/Salesforce. Its spiritual successor, Umbra, now serves as the foundation for research and teaching in high-performance data management at TUM. The group conducts cutting-edge research on various aspects of efficient, scalable, reliable, and secure systems software tailored for cloud environments.
Pramod Bhatotia, Prof. Dr.-Ing.
Alfons Kemper, Prof. Dr.
Gottfried Wilhelm Leibniz Prize: Umbra (Prof. Thomas Neumann)
The Gottfried Wilhelm Leibniz Prize is the most prestigious research award in Germany. Thomas Neumann was honored with this prize for his groundbreaking work on query compilation and high-performance database systems. The award supports research and development of Umbra, a high-performance data management system tailored for in-memory analytics, while also accommodating flash-based storage. Umbra achieves this by utilizing low-latency query compilation to machine code and fully leveraging the capabilities of modern hardware.
ERC Starting Grant DOS: A Decentralized Operating System (Prof. Pramod Bhatotia)
Advancements in computing hardware, AI, and big data have paved the way for numerous data-driven intelligent applications in fields such as health, mobility, and robotics. These applications rely on a vast number of decentralized computers to process user data. However, guaranteeing reliable and secure execution is challenging due to the need to program a diverse set of computers hosted across multiple untrusted jurisdictions. The Decentralized Operating System (DOS) project aims to address this issue through a groundbreaking hardware/OS co-design approach. In this model, reliability and security properties are enforced by the foundational layers of computing while adhering to policy requirements.
ERC Starting Grant CODAC: Commoditizing Data Analytics in the Cloud (Prof. Viktor Leis)
The landscape of data warehousing and analytics is shifting towards the cloud. However, current systems can be costly and frequently depend on proprietary storage formats, which effectively locks data in a single system. The objective of the CODAC project is to promote the commoditization of large-scale data analytics in public clouds. Commoditization entails making data analytics as affordable as possible while avoiding the vendor lock-in associated with any specific hyperscaler.
DFG Priority Program: Scalable Data Management on Modern Hardware
The objective of the project is to design and assess architectures and abstractions that enable flexible and scalable data management techniques. These techniques should offer extensibility for new data models, including processing and access mechanisms tailored for emerging applications. Additionally, they should capitalize on the features of modern and heterogeneous hardware, as well as operating system-level services.