Copenhagen, Denmark – November 19, 2025 — The Danish Centre for AI Innovation (DCAI) and WEKA, a leading AI-storage company, today announced the rollout of one of Europe’s first fully sovereign, end-to-end encrypted AI platforms. By integrating high-performance storage directly with the Gefion supercomputer, the collaboration provides European organisations with combined compute and storage infrastructure — all under full data sovereignty.
Unifying Compute and Storage for True AI Workflows
Until now, users of DCAI’s Gefion supercomputer primarily accessed compute resources, using separate systems for data storage. With this upgrade, Gefion gains 140 PB of usable storage, including 80 PB of NVMe SSDs optimised for storage performance via WEKA’s NeuralMesh™ software.
This integration allows customers to ingest data, train models, and deploy AI applications within a single, secure infrastructure, eliminating the need to shuttle large datasets between isolated storage and compute environments.
Nadia Carlsten, CEO of DCAI, commented:
“By adding storage to our platform, we’re making it easier for customers to run end-to-end AI workloads in one place — with the same performance, sovereignty, and security they rely on from our compute services.”
Performance That Matches Hyperscaler Scale
WEKA’s NeuralMesh architecture delivers sub-millisecond latency, ensuring high-frequency access to data and eliminating the I/O bottlenecks that typically throttle GPU utilisation. dcai.dk+1
By creating a zero-copy, high-throughput data path between storage and the Gefion GPU cluster, DCAI can deliver true multi-tenant orchestration — enabling research, enterprise, and government workloads to share infrastructure without compromising security or performance.
This architecture not only boosts GPU utilisation but also simplifies operational complexity and lowers TCO (Total Cost of Ownership) by avoiding data replication and cloud egress costs.
Efficiency, Sustainability & Cost Benefits
The integrated solution brings significant efficiency gains:
- A single-tier, software-defined architecture reduces the physical footprint compared to traditional multi-layer systems.
- Energy consumption is lower, since the need for data duplication or data transfer to external clouds is eliminated.
- The approach aligns with “green AI” objectives, enabling organisations to scale their AI infrastructure sustainably without sacrificing performance or sovereignty. dcai.dk
Building for Regulated and Sovereignty-Sensitive Use Cases
One of the most strategic benefits of this partnership is its suitability for highly regulated industries and sovereign use cases. Key features include:
- End-to-end encryption: No operator, including DCAI, has access to customer data, ensuring strong privacy and control.
- Full data sovereignty: The infrastructure aligns with European data regulation standards, addressing critical compliance needs.
- Independent audits and compliance: Designed to meet EU-mandated sovereignty policies, it offers assurances that enable organisations to innovate confidently. WEKA
Liran Zvibel, CEO and Co-Founder of WEKA, explained the broader vision:
“As the AI industry accelerates toward exascale, European innovators need infrastructure capable of scaling from petabytes to exabytes — without surrendering control. Our collaboration with DCAI delivers world-class performance with the data sovereignty required by modern regulations.”
Carlsten added:
“European organisations have long faced a tradeoff: either choose high-performance, hyperscaler infrastructure at the cost of data control, or retain sovereignty but settle for fragmented, slower systems. Now, that tradeoff is eliminated.” dcai.dk
Future Roadmap & Innovation Momentum
Looking ahead, DCAI and WEKA plan to continue building on this foundation:
- They will roll out WEKA’s Augmented Memory Grid™ to accelerate first-token response times for large language models and inference workloads.
- The platform aims to support dynamic bursting, allowing customers to scale GPU usage on demand for training or inference as workloads shift. WEKA
- Multi-tenant performance is already designed for consistency, with the infrastructure poised to sustain exabyte-scale growth without degrading performance. dcai.dk
Strategic Implications for Investors & Innovation Leaders
For Ixoraly’s audience — including investors, enterprise strategists, and AI infrastructure leaders — the implications of this announcement are extensive:
- Sovereign AI Market Validation: The deployment underlines demand in Europe for AI infrastructure that offers both high performance and data sovereignty.
- Next-Gen AI Factory: Organizations can now run full AI pipelines — data ingestion, training, inference — on a sovereign, high-performance platform.
- Green & Cost-Efficient Scaling: The single-tier architecture reduces infrastructure waste and energy use, aligning with ESG initiatives.
- Regulated Use Case Enablement: This system is tailor-made for regulated industries (public sector, defense, healthcare) that require secure, sovereign infrastructure.
- Strategic Infrastructure Positioning: By scaling to exabyte levels, DCAI + WEKA help future-proof Europe’s AI infrastructure for long-term growth.
Timeline & Availability
- Availability: The new storage-enabled Gefion infrastructure is now accessible to select enterprise, research, and government customers.
- Broader Onboarding: Wider customer onboarding is planned for Q1 2026, opening the platform to a broader set of European AI innovators.
- Future Enhancements: WEKA’s augmented memory and other advanced storage optimizations will be gradually deployed to further enhance performance and capacity.
About Ixoraly
Ixoraly is a forward-looking business intelligence platform delivering in-depth market analysis, foresight, and strategic insights at the intersection of AI, infrastructure, and global innovation. We support investors, corporate leaders, and technology teams by decoding how next-gen platforms transform industries and create new value.
