ScaDS Logo

COMPETENCE CENTER
FOR SCALABLE DATA SERVICES
AND SOLUTIONS

Thomas Neumann - HyPer: Building a Main-Memory Database System for Modern Hardware

Topic

The hardware trends of the last decade have led to a significant change in database architecture. Main memory sizes have grown dramatically, which implies that data is largely memory resident. As a consequence, CPU utilization becomes vastly more important as I/O wait times go away. In this talk, we revisit the development and evolution of the HyPer main-memory databases system. HyPer uses key techniques like query compilation and adaptive parallelization to utilize modern hardware, and achieves excellent performance over a wide range of uses cases.

Short Bio

Thomas Neumann is a full professor for Data Science and Engineering at the Technical University of Munich. His researches focus on query optimization and query processing in large databases. He has build several successful systems, including the RDF-3X engine for graph data and the HyPer database management system for relational data. He was awarded the VLDB Early Career Research Contribution Award for his work in query processing and optimization.