SIMPSON4NFDI (NFDI-Storage 2025)

• Term: Funded in 2026
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 579652141

We aim to set up NFDI-storage at the Leibniz Supercomputing Center (LRZ), the academic IT provider of LMU and TUM and one of three top-tier supercomputing centers in Germany. The Munich Universities and LRZ have a strong record of contributions to the NFDI landscape and intend to create an environment where consolidated and sustainable NFDI services can be optimally supported. Five consortia (PUNCH4NFDI, NFDI4Ing, DAPHNE4NFDI, FAIRmat and NFDI4Earth) are supporting this proposal: PUNCH4NFDI, the consortium of particle, astroparticle, astro-, hadron and nuclear physics, covers a substantial fraction of basic research in physics, and in particular of data-intense physics at large facilities. NFDI4Ing covers the breadth of the engineering sciences catering RDM solutions for a variety of use cases from large numbers of small data sets to huge HPC and measurement data sets. DAPHNE4NFDI serves the broad community of users employ-ing a wide range of photon and neutron techniques. The community performs thousands of individual user experiments every year. FAIRmat covers the full breadth of the condensed matter and materials physics supplying FAIR solutions and a federated data hub for materials data. NFDI4Earth finally addresses digital needs of Earth System Sciences cooperating in international and interdisciplinary networks with the overarching aim to understand the func-tioning and interactions within the earth system and to tackle the multiple challenges of global change. The proposed system can accommodate hot and warm data from measurements, High-Performance Computing (HPC), Big Data reprocessing and Artificial-Intelligence (AI) efforts. Due to its reliability and connectivity, the proposed system will be ideal for serving lukewarm or cold data to a broad community as well – enabling data publication and FAIR Research Data Management (RDM). Thus, the system bridges environments and communities creating huge and valuable datasets with the idea of FAIR RDM and NFDI. We plan for 27.5 PB of storage at LRZ as Data Science Storage (DSS) building blocks. The storage will be integrated into the well-established LRZ-DSS management, taking advantage of a well-established user-access and data-sharing scheme. The system can be used from the regional- and national-scope HPC clusters of LRZ, and also via virtual machines running on a dedicated (~500 CPU core) Cloud-Computing Frontend component, which we propose for storage interfacing. Our concept allows NFDI consortia to use virtual machines to run enhanced interfaces for FAIR RDM (repository frontends, storage middleware, data management and access frame-works). Mass-data transfer tools avaliable at LRZ (GLOBUS, GridFTP, UFTP) support high-throughput data sharing over the German Research Network (DFN) and the internet. The proposed system design enables excellent IaaS, PaaS and SaaS (Infrastructure-/Platform-/Software-as-a-Service) solutions for FAIR data management by the NFDI and its consortia.

• DFG Programme: Major Research Instrumentation
• Major Instrumentation: SIMPSON4NFDI (NFDI-Speicher 2025)
• Instrumentation Group: 7000 Datenverarbeitungsanlagen, zentrale Rechenanlagen
• Applicant Institution: Ludwig-Maximilians-Universität München
• Leader: Professor Dr. Joseph Mohr