One of the most ambitious efforts comes from the Jülich Brain Atlas project, which recently developed CytoNet, a foundation model designed to analyze brain microarchitecture. Led by neuroscientist Katrin Amunts and computer scientist Christian Schiffer, the project uses cellular-scale brain imaging data to connect individual cell structures with broader patterns of brain organization and function.
Researchers trained the model on 6.5 petabytes of data collected from 21 post-mortem brains, completing the process in less than five days using 4,096 NVIDIA Grace Hopper Superchips. The team is now working toward an AI agent for brain researchers that combines multimodal reasoning, language interfaces, and question-answering capabilities.
“For the first time, we’re not just using AI to analyze the brain — we’re building an agent that can think through the experiment itself,” said Katrin Amunts, director of INM-1 at Forschungszentrum Jülich and professor of brain research at Heinrich Heine University Düsseldorf. “That changes what neuroscience will be, and JUPITER is what makes that sentence possible to say today.”
Climate science is another area where JUPITER is being deployed at scale. Researchers from ETH Zurich, the German Climate Computing Centre, Jülich Supercomputing Centre, the Max Planck Institute for Meteorology, NVIDIA, the Swiss National Supercomputing Centre, and the University of Hamburg developed a new ICON configuration that received the Gordon Bell Prize for Climate Modelling at SC25.
The system is described as the first model capable of simulating a fully coupled Earth system at 1-kilometer resolution, incorporating interactions between the atmosphere, oceans, land systems, biogeochemistry, and the carbon cycle. Running on 20,480 NVIDIA Grace Hopper Superchips, the model simulated approximately 146 days of real-world climate conditions in 24 hours of computing time.
“Our simulations resolve the fine-scale winds, ocean eddies and upper-ocean mixing that shape marine ecosystems and regulate the ocean’s uptake of carbon,” said Daniel Klocke, computational infrastructure and model development group leader at the Max Planck Institute for Meteorology. “At a global resolution of just 1 kilometer, many of these interactions emerge directly from the laws of physics rather than being approximated. This gives us an unprecedented view of how the atmosphere, ocean and biosphere work together, helping us understand the processes driving our changing climate.”
JUPITER is also being used in telecommunications research. Ericsson and Forschungszentrum Jülich recently announced a collaboration focused on developing AI technologies for the continued evolution of 5G networks and future 6G systems. The supercomputer will serve as the platform for large-scale AI model training and testing.
The research effort is centered on brain-inspired computing architectures intended to support complex network operations while reducing energy consumption. Areas of study include AI models for radio and core networks, neuromorphic approaches for edge inference, and modular supercomputing concepts derived from exascale computing research.
In quantum computing, researchers at the Jülich Supercomputing Centre and the NVIDIA Application Lab achieved a new milestone by fully simulating a universal 50-qubit quantum computer. The result surpassed the previous 48-qubit benchmark and was enabled by JUPITER’s tightly integrated CPU-GPU memory architecture, which allows larger quantum states to be processed than would fit within GPU memory alone.
The resulting simulator, known as JUQCS-50, will be made available through JUNIQ, JSC’s quantum computing user facility. Researchers plan to use the platform to evaluate quantum algorithms and explore future quantum computing architectures.
Taken together, the projects demonstrate how JUPITER is being applied across a diverse range of scientific and engineering challenges. From mapping brain structures and modeling climate systems to advancing wireless communications and quantum computing research, the exascale system is serving as a platform for workloads that demand unprecedented computing scale.
This analysis is based on reporting from NVIDIA.
Image courtesy of NVIDIA.
This article was generated with AI assistance and reviewed for accuracy and quality.
