TLDR: NVIDIA has announced a significant collaboration with leading model builders and cloud providers across Europe and the Middle East. This initiative aims to optimize sovereign large language models (LLMs) tailored to local languages and cultures, thereby accelerating enterprise AI adoption. Leveraging NVIDIA’s Nemotron techniques, the partnership seeks to enhance cost efficiency and accuracy, with optimized models hosted on regional AI infrastructure and integrated into platforms like Perplexity.
Paris, France – NVIDIA, a global leader in accelerated computing, has unveiled a strategic collaboration with a consortium of prominent model builders and cloud providers spanning Europe and the Middle East. The ambitious initiative, announced at GTC Paris during VivaTech, is designed to optimize sovereign large language models (LLMs), providing a robust foundation to accelerate enterprise AI adoption across the region’s diverse industries.
The core objective of this partnership is to develop AI models that are deeply tailored to local languages and cultures, ensuring greater relevance and effectiveness for businesses operating within these regions. NVIDIA’s advanced Nemotron techniques are central to this effort, promising to maximize cost efficiency and accuracy for a wide array of enterprise AI workloads, including the burgeoning field of agentic AI. These optimization processes involve sophisticated methods such as neural architecture search, reinforcement learning, and post-training with NVIDIA-curated synthetic data, all aimed at reducing operational costs and boosting user experiences by generating tokens faster during inference.
Key participants in this collaborative endeavor include a diverse group of model builders and AI consortiums: Barcelona Supercomputing Center (BSC) from Spain, Bielik.AI from Poland, Dicta from Israel, H Company and LightOn from France, Domyn from Italy, the National Academic Infrastructure for Supercomputing in Sweden (NAISS) alongside KBLab at the National Library of Sweden, the Slovak Republic, the Technology Innovation Institute (TII) from the UAE, the University College London from the UK, the University of Ljubljana, and UTTER. These entities are working directly with NVIDIA to refine their models using the Nemotron framework.
Once optimized, these models will be hosted on cutting-edge AI infrastructure provided by NVIDIA Cloud Partners (NCPs) within Europe. Notable NCPs participating in the NVIDIA DGX Cloud Lepton marketplace include Nebius, Nscale, and Fluidstack. This regional hosting ensures data sovereignty and compliance with local regulations, a critical factor for many European enterprises.
A significant aspect of this collaboration is the integration of these open, sovereign models with Perplexity, an AI-powered answer engine that currently processes over 150 million questions per week. This integration will enhance the accuracy of search queries and AI outputs for European enterprises. Furthermore, companies will have the capability to fine-tune these sovereign models on local NCP infrastructure through a new integration with Hugging Face and DGX Cloud Lepton, offering unparalleled flexibility and customization.
Jensen Huang, founder and CEO of NVIDIA, underscored the significance of this initiative, stating, “Europe’s diversity is its superpower — an engine of creativity and innovation. Together with Europe’s model builders and cloud providers, we’re building an AI ecosystem where intelligence is developed and served locally to provide a foundation for Europe to thrive in the age of AI — transforming every industry across the region.”
Also Read:
- Nvidia-Backed CoreWeave’s Explosive Growth Disrupts AI Cloud Market
- Ireland’s Collaborative Strategy Aims to Solidify its Position as a Global AI Hub
The partnership explicitly supports Europe’s 24 official languages, aiming to accelerate the region’s AI-driven transformation across vital sectors such as manufacturing, robotics, healthcare, pharmaceuticals, finance, energy, and creative industries. The first models resulting from this extensive collaboration are anticipated to become available later in the year, marking a pivotal step towards a more localized and efficient AI landscape in Europe and the Middle East.
