Ericsson Partners with Forschungszentrum Jülich to Advance AI-Driven 6G
Ericsson and Forschungszentrum Jülich have signed a Memorandum of Understanding to jointly develop advanced AI technologies for next-generation telecom networks, targeting both the evolution of 5G and the foundation of 6G. The collaboration will leverage Europe’s first exascale supercomputer, JUPITER, operated by the Jülich Supercomputing Centre, to design and validate AI models capable of addressing the scale and complexity of future mobile networks.
The joint effort focuses on integrating high-performance computing (HPC) with telecom system design, enabling more efficient network architectures across core, RAN, and network management domains. The partners will investigate AI-driven optimization techniques, large-scale model training, and advanced benchmarking methodologies to evaluate performance, scalability, and efficiency. Initial commercial 6G deployments are expected around 2030, aligning with broader industry timelines.
A central research theme involves neuromorphic computing approaches, including memristor-based architectures, aimed at significantly improving energy efficiency for AI inference at the network edge. The collaboration will also explore Modular Supercomputing Architecture (MSA) concepts and operational strategies such as heat reuse to improve sustainability across both cloud and HPC environments. The initiative aligns with Europe’s broader EuroHPC efforts to build sovereign, high-performance digital infrastructure.
- Collaboration targets AI-driven architectures for 5G evolution and future 6G networks
- Utilizes JUPITER, Europe’s first exascale supercomputer, for large-scale AI model development
- Focus areas include core network, RAN, and network management optimization
- Neuromorphic computing (e.g., memristors) explored for energy-efficient AI inference at the edge
- Research includes benchmarking AI models for scalability, storage efficiency, and execution speed
- Investigates Modular Supercomputing Architecture (MSA) and heat recovery for sustainable HPC/cloud operations
- Supports Europe’s EuroHPC initiative to strengthen sovereign digital infrastructure
Nicole Dinion, Head of Architecture and Technology, Cloud Software and Services at Ericsson, said: “The future of mobile networks is deeply intertwined with AI and the need for unparalleled energy efficiency. Our collaboration with Forschungszentrum Jülich combines their research and computing power with our expertise in telecoms technology to explore architectures that define the next generation of telecommunication.”
🌐 Analysis
This collaboration positions Ericsson within a growing cohort of telecom and infrastructure vendors aligning with HPC ecosystems to accelerate AI-native network design. Companies such as Nokia and NVIDIA have also expanded partnerships around AI-driven RAN, digital twins, and accelerated computing platforms, reflecting a broader convergence between telecom and supercomputing domains. Ericsson’s focus on neuromorphic computing adds a differentiated research vector, especially as energy efficiency becomes a limiting factor in scaling AI-driven networks.
The use of exascale infrastructure such as JUPITER underscores a shift toward simulation-heavy, AI-assisted network engineering workflows. This mirrors trends seen in hyperscale environments where large-scale model training and system optimization increasingly rely on HPC-class resources. As 6G research accelerates globally—with initiatives across Europe, the U.S., and Asia—partnerships that integrate telecom expertise with advanced computing platforms are likely to define early architectural directions for next-generation networks.