What are JUPITER's specific technical specifications and how does it support 6G network R&D?

JUPITER is Europe's first exascale supercomputer with peak performance exceeding 1 Exaflop (FP64 precision) and AI inference capability reaching 93 Exaflops (FP8 precision). It features approximately 23,762 NVIDIA GH200 Grace Hopper Superchips and over 1.5 PB of HBM3 high-bandwidth memory. It supports 6G R&D by enabling research teams to simulate complex wireless communication environments at unprecedented scales, train spectrum management AI systems, develop brain-inspired computing algorithms, and simulate large-scale distributed learning scenarios.

What are the specific application directions for brain-inspired computing in 6G networks?

Brain-inspired computing has three main application directions in 6G: First, Spiking Neural Networks (SNNs) that can complete beam management and user scheduling at millisecond timescales while consuming only one-hundredth of traditional GPU inference energy. Second, neuromorphic chip architectures that process network data in an event-driven manner, significantly reducing static power consumption for base stations. Third, federated learning and distributed intelligence that leverages JUPITER's parallel computing to research how thousands of base stations can collaboratively optimize network parameters while preserving user privacy.

What is Europe's competitive advantage and strategic positioning in the global 6G race?

Europe's advantages in the 6G race lie in its deep tradition of fundamental research and strong industrial capabilities. Ericsson and Nokia hold core wireless communication technology patents, while the EuroHPC initiative provides world-class supercomputing infrastructure. The EU's Horizon Europe framework has allocated over €900 million for 6G research, with the flagship Hexa-X-II project led by Ericsson bringing together resources from 25 European countries. JUPITER's involvement further strengthens Europe's computational support for 6G-AI convergence research.

Ericsson and Forschungszentrum Jülich to Develop Advanced AI for 6G Networks Using Europe's First Exascale Supercomputer

Ericsson and Jülich Research Centre's Strategic Partnership: Exascale Computing Powers 6G and AI Convergence

Introduction: A Historic Convergence of Supercomputing and Telecommunications

Ericsson and Jülich Research Centre's Strategic Partnership: Exascale Computing Powers 6G and AI Convergence

Introduction: A Historic Convergence of Supercomputing and Telecommunications

In March 2026, Swedish telecommunications giant Ericsson announced a strategic partnership with Germany's Forschungszentrum Jülich (Jülich Research Centre) to leverage JUPITER, Europe's first exascale supercomputer, for the joint development of artificial intelligence technologies targeting 6G networks. This collaboration marks a profound convergence of supercomputing and telecom network research, signaling that next-generation mobile communications will fundamentally depend on AI-driven innovation.

JUPITER, which became operational at the Jülich Supercomputing Centre in late 2024, delivers peak performance exceeding 1 Exaflop (10^18 floating-point operations per second), making it Europe's most powerful supercomputer. Built on NVIDIA's Grace Hopper Superchip architecture with approximately 24,000 GPU accelerators, the system is purpose-built for large-scale AI training and scientific simulation. Ericsson's decision to partner with JUPITER reflects the unique advantages the system offers in processing massive telecommunications network data and training complex AI models.

JUPITER's Technical Specifications and Capabilities

JUPITER is one of the flagship projects of the EuroHPC Joint Undertaking, with total investment exceeding €500 million. Its technical specifications are remarkable:

**Peak Performance**: Over 1 Exaflop (FP64 precision), with AI inference capability reaching 93 Exaflops (FP8 precision)
**GPU Count**: Approximately 23,762 NVIDIA GH200 Grace Hopper Superchips
**Total Memory**: Over 1.5 PB of HBM3 high-bandwidth memory
**Storage System**: Tiered storage architecture with total capacity exceeding 100 PB
**Network Interconnect**: Based on NVIDIA Quantum-2 InfiniBand network with 400 Gb/s bandwidth
**Energy Efficiency**: Warm-water cooling system improving energy efficiency by over 40% compared to conventional supercomputers

These hardware specifications make JUPITER an ideal platform for training ultra-large-scale AI models. For 6G network research, this means research teams can simulate complex wireless communication environments and train AI systems capable of real-time network performance optimization at unprecedented scales.

Brain-Inspired Computing and Communication Network Intelligence

One of the core research directions of this partnership is "Brain-Inspired Computing," which draws upon the neural network structure and information processing mechanisms of the human brain to develop novel AI algorithms and computing architectures. Ericsson's research team believes that traditional deep learning approaches face limitations in computational latency and energy consumption when handling real-time optimization problems in communication networks, and brain-inspired computing promises to break through these bottlenecks.

Specific application directions of brain-inspired computing in 6G networks include:

Spiking Neural Networks (SNNs): These networks simulate the pulse signal transmission mechanisms of biological neurons, enabling ultra-low-power edge intelligent inference. In base station deployments, SNNs can complete beam management and user scheduling decisions at millisecond timescales while consuming only one-hundredth of the energy required by traditional GPU inference.

Neuromorphic Chip Architectures: Development of specialized communication AI chips that bring neuromorphic computing to base station and core network equipment. These chips process network data in an event-driven manner, significantly reducing static power consumption and making them ideal for always-on network monitoring and optimization scenarios.

Federated Learning and Distributed Intelligence: Leveraging JUPITER's parallel computing capabilities to simulate large-scale distributed learning scenarios, researching how thousands of base stations can collaboratively learn and optimize network parameters while preserving user privacy.

The AI-Native Architecture Vision for 6G Networks

6G networks are expected to begin commercial deployment around 2030, and the industry widely positions them as "AI-Native Networks." Unlike the 5G era where AI primarily served as an auxiliary tool, 6G's core architecture will deeply integrate AI capabilities from the design stage. The Ericsson-Jülich partnership is precisely laying the technical foundation for this vision.

In spectrum allocation, 6G will face unprecedented spectrum management challenges. With the introduction of terahertz (THz) frequency bands, available spectrum resources will increase dramatically, but signal propagation characteristics will also become more complex. AI-driven dynamic spectrum allocation systems need to sense electromagnetic environments in real time, predict user demands, and coordinate spectrum sharing among different operators and services. JUPITER provides the necessary computational resources for training such complex spectrum management AI systems.

Beam management is another critical application domain. 6G networks are expected to extensively employ ultra-massive MIMO antenna arrays, with beam counts reaching hundreds or even thousands. Traditional beam search and tracking algorithms become extremely inefficient at such scales, while AI-based beam management solutions can achieve predictive beam switching and adaptive beamforming by learning user mobility patterns and channel characteristics.

Network slicing technology will also undergo a qualitative leap in the 6G era. Current 5G network slicing is primarily based on static policy configurations, whereas 6G will achieve fully autonomous intelligent slice management. AI systems will need to dynamically create, adjust, and release network slice resources based on real-time service demands, network load conditions, and quality-of-service requirements.

Europe's Strategic Position in the 6G Race

This partnership also reflects Europe's strategic considerations in the global 6G technology race. Currently, the United States, China, South Korea, and Japan are all accelerating their 6G research and development. The US Next G Alliance, China's IMT-2030 Promotion Group, South Korea's 6G research programs, and Japan's Beyond 5G Promotion Consortium have all committed substantial resources.

Europe's advantage lies in its deep tradition of fundamental research and strong industrial capabilities. Ericsson and Nokia, as two of the world's major telecommunications equipment suppliers, hold core wireless communication technology patents. The EuroHPC initiative provides Europe with world-class supercomputing infrastructure. Combining both creates a unique competitive advantage in the 6G-AI convergence domain.

The EU's Horizon Europe framework programme has allocated over €900 million for 6G research, with flagship projects like "Hexa-X-II" led by Ericsson, bringing together research institutions and enterprises from 25 European countries. JUPITER's involvement provides unprecedented computational support for these projects and is expected to accelerate key technological breakthroughs.

Industry Impact and Market Outlook

From an industry perspective, the convergence of supercomputing and telecommunications will spawn new business models and industrial ecosystems. Future telecom operators may need to build their own AI training infrastructure or access supercomputing capabilities through cloud services to continuously optimize network performance. This will drive deep integration between the telecom and high-performance computing industries.

According to market research projections, the global 6G market is expected to reach several trillion dollars between 2030 and 2040. AI-driven network optimization and intelligent services will account for a significant share. Through this partnership, Ericsson is positioning itself ahead of the curve, poised to capture a first-mover advantage in this emerging market.

Conclusion

The partnership between Ericsson and the Jülich Research Centre transcends a simple collaboration between a corporation and a research institution. It represents a historic convergence of three frontier domains: supercomputing, artificial intelligence, and next-generation communications technology. JUPITER's formidable computational power will serve as an accelerator for 6G-AI technological breakthroughs, while cutting-edge research directions such as brain-inspired computing promise to fundamentally transform the intelligence level of communication networks. As the global 6G race intensifies, this partnership secures a valuable technological head start for Europe.