HPI Data Center

Computing power for research and teaching: The HPI Data Center, situated on “Campus II,” is our in-house computing center. The building is committed to pooling our computational capacities.

Construction of the center, which was finished in 2022, is part of our ongoing growth and mission of strengthening Europe's digital sovereignty.

For more sustainability

The thermal discharge of the server infrastructure is used for the heating supply of the entire building.

Portrait of Head of Data Engineering Systems Prof. Dr. Tilmann Rabl

The data center is vital for the research at HPI. It provides cost and energy efficient compute and enables research on all levels of the infrastructure, well beyond the possibilities of public clouds. Direct access to the resources is a significant advantage for HPI, our students, faculty and researchers.

Prof. Tilmann Rabl, Head of Data Center / Head of Data Engineering Systems

About the Data Center

Our computing center offers decisive advantages for our researchers and their research work. For example, we do not have to rely on computing power from the cloud for our research, thus creating independent research opportunities. Direct access and the associated possibilities for intervention and modularity underline the advantages.
The Data Center is not just a centralized resource but also a flexible platform that can accommodate a variety of solutions tailored to the needs of our departments and research groups.
A broad spectrum of hardware and software architectures, ranging from “legacy systems” to the most modern technology, allows for various research projects and approaches. This spectrum reflects the real situation in data centers, especially in security research, making research results even more scalable.

Our data center acts both as a site for research and as a subject of research.

Research site: The HPI Data Center enables excellent research and is also a research object. The increasing demand for computational power due to AI and the ongoing digitalization requires a rethinking of resource utilization.
Subject of research: How computing power can be used more effectively, efficiently and thus sustainably and how data centers can be designed to meet these requirements is the goal and focus of current HPI research.

Data Engineering Lab

Our Data Engineering Lab is part of the HPI Data Center and an experimental platform for scientists from the HPI departments.
The lab creates unique research opportunities for machine learning and data analysis research.
The emphasis is on methods for processing and storing large and dynamic data sets and on complex data analysis on parallel distributed infrastructures and modern database system architectures.

The Data Engineering Lab provides the research projects at HPI with a wide range of robust and modern computer architectures.
This includes high-performance servers with extensive main memory and powerful processors, computer clusters with many nodes and cores, GPU servers with many computationally optimized graphics cards, modern hardware systems with, for example, non-volatile RAM modules and power architectures, and SAN systems with sufficient storage capacity and high read/write speeds.
The lab's hardware landscape is designed for expandability so that the research lab can grow with a growing demand for project capacities at HPI, and new hardware trends can be followed and realized in the lab.

On top of the powerful hardware, the Lab offers various state-of-the-art software systems for processing, transforming, analyzing, storing, and machine learning data.
This off-the-shelf software includes runtime environments for major programming languages, frameworks for machine learning and distributed computing, and selected proprietary analysis software and databases.
The Lab is serving as a repository for research and industry project data that will be processed and analyzed in the Lab.

AI Service Center

The AI Service Center Berlin-Brandenburg at HPI, funded by the Federal Ministry of Education and Research is your gateway to AI. We empower startups, companies, and institutions to embrace AI through free resources, training, and expert advice.

Find out more