Deep Learning Cluster for Retrieval-Augmented Generation and Provenance Analytics

• Subject Area: Computer Science
• Term: since 2026
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 583682641

In order to ensure the competitiveness and innovative capacity of AI research at the Bauhaus University Weimar, a new deep learning cluster is needed that is designed to meet the increased requirements of current deep learning models. The main applicant, the research group "Intelligent Information Systems", is part of the Media Informatics Department of the Faculty of Media at the Bauhaus-Universität Weimar. The working group has experience in operating compute clusters for large-scale storage and processing of web data, as well as for deep learning research in the areas of Information Retrieval and Natural Language Processing (NLP). In addition to its internationally recognized research, the group is known for its community services, such as publicly accessible research search engines and the organization of shared tasks. The planned deep learning cluster will also be available to other computer science research groups - in particular the newly appointed Computational Social Science group - as well as computer science-related groups in the Faculty of Civil Engineering, and project partners from other universities. It will thus enable innovative AI research across disciplines and strengthen collaboration at the university and beyond through interdisciplinary projectss. With regard to the further development of the HS-ITZ, Prof. Matthias Hagen (Friedrich Schiller University Jena), a research partner in several of the above-mentioned projects, was closely involved in the configuration of the planned acquisition. In terms of space and organization, the hardware purchase has been optimally prepared: With the renovation of the computer science faculty building completed in 2024, a state-of-the-art computer room for such clusters was built and put into operation. It offers sufficient capacity for the expansion of the deep learning infrastructure; support and maintenance of the systems are ensured by the existing specialist staff.

• DFG Programme: Major Research Instrumentation
• Major Instrumentation: Deep-Learning-Cluster
• Instrumentation Group: 7040 Vektorrechner
• Applicant Institution: Bauhaus-Universität Weimar
• Leader: Professor Dr. Benno Stein