The InHPC-DE project will further integrate the three German national HPC sites that comprise the Gauss Centre for Supercomputing (GCS). The overarching goal of the project is to create a national, virtual center for supercomputing that enables the consistent technical and organizational use of HPC services. In particular, this includes evaluating the new requirements arising from possible support and integration with the Gaia-X ecosystem. For this purpose, a GCS Gaia-X laboratory will be established, which will be operated in part at each of the three locations — HLRS along with the Leibniz Supercomputing Centre and the Jülich Supercomputing Centre.
The goal of the research project is to develop the necessary processes and tools to enable the efficient management of user data, the uniform control of simulation calculations and their evaluation across the three sites as transparently as possible for users from science and industry.
In the project InHPC-DE, various services will be analyzed, implemented and tested to enable greater technical integration of the different world-class supercomputers. The aim is to investigate and test how future exascale applications can be optimally supported via tightly coupled, virtual infrastructures and what additional performance increase is possible through the simultaneous use of the GCS systems.
InHPC-DE is a follow-up proposal, with an increased focus on developing new scenarios for the GCS systems, gathering user feedback, and supporting pilot users. The addition of the security and Gaia-X work packages to the project has become necessary to fulfill the generally required components (network, data management, federated access and compute) of a Gaia-X node.
01. January 2022 - 31. December 2023
Artificial Intelligence & Data Analytics
Cloud Computing
Exascale Computing
HPC for Industry
HPC Infrastructure & Networks
Visualization
High-Performance Computing Network - Production
Project and User Management, Accounting
Converged Computing
BMBF
See all projects
Head, High-Performance Computing Network - Production