DataCrunch is positioning itself as a major European competitor to US hyperscalers by offering a sovereign, scalable, and sustainable AI cloud infrastructure rooted entirely in Europe.

The company is building a gigafactory in Latvia with around 100,000 AI accelerators, aiming to reduce dependency on dominant US cloud giants like Amazon, Google, and Microsoft, which control nearly 70% of the European cloud market.

All operations in Finland and Iceland are powered by 100% renewable energy, emphasizing DataCrunch's commitment to sustainability and environmentally friendly practices.

Founded in 2018 by Ruben Bryon, DataCrunch has raised €55 million in a Series A funding round to expand its high-performance AI infrastructure, with plans to deploy NVIDIA Blackwell-based systems and open a new facility in Akaa, Finland.

Since its founding, the company has secured a total of €76.5 million in funding from investors including byFounders, Skaala, Varma pension fund, Tesi, and debt financing from Nordea, Armada Credit Partners, Danske Bank, Norion Bank, and LocalTapiola.

DataCrunch has partnered with notable AI startups like 1X and Unbabel, as well as companies such as Sony and Freepik, launching new services including on-demand NVIDIA HGX B200 access, Instant Clusters with InfiniBand, and Serverless Containers.

The company has introduced innovative products like Instant Clusters for distributed AI workloads and Serverless Containers capable of handling hundreds of millions of requests monthly, with future plans to enhance features such as Managed Kubernetes and distributed Object Storage.

DataCrunch has proposed establishing a European AI GigaFactory in Latvia, which could house up to 100,000 accelerators for large-scale model training and inference, supported by the European Commission and Latvian partners.

The company plans to deploy advanced GPUs, including Nvidia’s B300 and GB30 systems, to bolster its infrastructure and provide secure, regulation-compliant AI compute capacity across the EU.

Future platform enhancements include managed Kubernetes, distributed object storage, custom inference endpoints, and improvements in IAM, APIs, and network automation to support AI development and deployment.

DataCrunch specializes in high-performance, cost-effective GPU solutions for AI training and inference, serving clients like Sony, Freepik, Schibsted, and Unbabel, and is among the first globally to deploy NVIDIA’s HGX B200 systems.

The company is increasingly attracting European customers who prioritize security and sovereignty, seeking local alternatives to US-based cloud providers.