Multiverse Computing, YellowDog target European sovereign AI

Author: Joe Peck

Multiverse Computing, a Spanish company specialising in compressing large language models (LLMs), and UK cloud workload orchestration firm YellowDog.ai have agreed to work together to improve compute efficiency and support the development and deployment of European sovereign AI.

The collaboration will combine AI model compression with hybrid and multi-cloud compute management to help enterprises, research organisations, and public sector institutions operate AI within European infrastructure.

Multiverse Computing’s CompactifAI technology uses quantum-inspired tensor-network methods to compress large AI models, while YellowDog.ai provides a platform for managing compute resources across hybrid and multi-cloud environments.

The companies say they aim to bring together processing capacity, secure access, storage, networking, and data centre operations, whilst simultaneously reducing costs, energy consumption, and infrastructure complexity. They also identify GPU shortages as a key challenge, particularly for European organisations seeking to make more efficient use of available capacity.

Customers will be able to use the combined capabilities for model development, benchmarking, optimisation, compressed model deployment, inference, and sovereign infrastructure planning. Target sectors include financial services, life sciences, advanced engineering, energy, telecoms, public services, and defence-adjacent industries, where data residency, security, cost, compute availability, and regulatory compliance are important considerations.

Supporting Europe’s AI compute infrastructure

Enrique Lizaso, co-founder and CEO of Multiverse Computing, says, “Sovereign AI will only succeed if it is economically and technically sustainable. Our compressed AI models are designed to run faster, use less energy, and operate closer to the data.

“Working with YellowDog.ai gives customers a powerful route to deploy efficient AI across cloud, on-premises, edge, and sovereign data centre environments.”

Bruce Beckloff, CEO of YellowDog.ai, adds, “Europe has the talent, data, and industrial base to lead in AI, but it needs infrastructure that is fast, efficient, secure, and sovereign by design.

“By partnering with Multiverse Computing, we are creating practical capacity solutions where organisations can scale compute intelligently, reduce bottlenecks, and retain control of critical workloads.”

The agreement comes as European organisations seek alternatives to fragmented, expensive, and capacity-constrained AI infrastructure. The companies also point to the European Commission’s AI Gigafactory initiative, which envisages large-scale facilities providing computing capacity for the development, training, fine-tuning, and deployment of advanced AI models and applications.

These facilities are expected to combine AI-optimised computing, secure cloud access, high-capacity storage, advanced networking, energy-efficient data centres, and specialist support services.

The stated priorities include improving access to hybrid compute, reducing operational costs and energy use through model compression and workload scheduling, and supporting deployments that meet European requirements for data protection, security, and technological autonomy.

For more from Multiverse Computing, click here.