Nvidia drives regional sovereign data infrastructure with new partnership

Nvidia logo on the side of a building
(Image credit: Shutterstock)

Nvidia and cloud service provider Scaleway have announced a new partnership aimed at driving the availability of sovereign infrastructure across Europe. 

The two have teamed up to offer European startups cloud credits, access to GPUs, Nvidia AI Enterprise software, and large language models (LLMs) all on sovereign infrastructure based in the EU.

Through the deal, Scaleway, a subsidiary of French telecoms firm Iliad, is offering cloud credits for access to its AI supercomputer cluster, with 1,016 Nvidia H100 Tensor Core GPUs.

The firm said its sovereign infrastructure push aims to ensure compliance with EU data protection legislation. The firm’s data centers hold compliance certifications that ensure data security for critical services such as healthcare and public services.

This expanded access to sovereign computing comes via Nvidia's Inception program, a free global scheme that provides technical guidance, training, discounts, and networking opportunities. 

It's already attracted startups such as Hugging Face, which hosts operations in both the US and France.

"AI is the new way of building technology, and making the fastest AI accelerators accessible within regional clouds is key to democratizing AI across the world, enabling enterprises and startups to build the experiences of tomorrow," said Jeff Boudier, head of product at Hugging Face.

Scaleway said its new Nabuchodonosor supercomputer - an Nvidia DGX SuperPOD with 127 Nvidia DGX H100 systems - will help startups in France and across Europe scale up AI workloads.


Whitepaper cover with title over an image of a city with a lightning bolt shaped cloud above in the blue sky

(Image credit: CDW)

Looking to get more from your cloud experience? Discover the benefits of multi-cloud designed by Dell and CDW.


Regional Inception members will also be able to access Nvidia AI Enterprise software on Scaleway Marketplace, including the Nvidia NeMo framework and pretrained models for building large language models. 

Access to Nvidia Rapids for accelerated data science will also be offered alongside options for Nvidia Triton Inference Server and Nvidia TensorRT-LLM for boosting inference.

This will offer access to cloud computing credits, Nvidia Deep Learning Institute courses, technology experts, preferred pricing on hardware and software, and guidance on the latest software development kits and AI frameworks.

 Heightened data sovereignty concerns 

Under EU regulations, cloud services must be operated and personal data held within the EU to ensure data security and privacy standards. This practice has been introduced due to lingering concerns that data could be shared with US government agencies. 

Data sovereignty has become a regulatory flashpoint in recent months, with proposed new rules potentially requiring non-EU cloud providers such as Microsoft, Google, and AWS to enter into partnership with EU firms if they want to operate within the union.

Serge Lemonde, Nvidia Inception program lead for EMEA and India, said the partnership will help shore up data sovereignty efforts for participating organizations.

"When doing business in Europe, US companies, for example, need to comply with EU regulations on sovereignty to secure data against access from foreign adversaries or entities," he said. 

"Noncompliance risks data vulnerabilities, financial penalties and legal consequences."

Emma Woollacott

Emma Woollacott is a freelance journalist writing for publications including the BBC, Private Eye, Forbes, Raconteur and specialist technology titles.