NTT Deploys Nvidia Blackwell GPU Test Cluster at Munich 2 Data Center

NTT Global Data Centres has deployed an Nvidia GPU cluster at its Munich 2 data center, giving enterprise customers a production-grade environment to test AI applications on Nvidia Blackwell hardware before committing to large-scale deployments of their own.

The cluster, branded AI Factory Munich, is designed to go beyond the typical AI demonstration or innovation lab. It allows customers to evaluate AI workloads against the infrastructure, security, data governance and operational requirements they will ultimately face in production, according to the company. The facility is built on Nvidia Blackwell GPUs interconnected via Nvidia Spectrum-X Ethernet technology, with servers and storage supplied through Dell's AI Factory offering.

Anne-Sophie Lotgering, CEO Europe at NTT, said in a LinkedIn post that Nvidia and Dell provided the infrastructure, while Red Hat and Mistral AI collaborated with the company on the platform and the model. The involvement of Red Hat and Mistral AI underscores a broader industry push toward integrating open-source software stacks and European-built AI models into enterprise-grade infrastructure.

Graham Clarke, managing director of global private and sovereign AI at NTT Data, framed the initiative as a way for organizations to retain control over their AI deployments. "Every leader I speak to wants AI they own, control and trust. At NTT Data's AI factory Munich, they can run their own applications and data in a production data center, on their terms, before they commit to deployment. It's the difference between a slide and a decision," he said.

The Munich 2 campus is located in the German town of Unterschleißheim outside Munich, Bavaria. It spans 6,300 square meters (67,813 square feet) across two buildings and offers 14MW of critical IT load. The deployment reflects a growing trend among data center operators to move beyond raw colocation and power provision by offering managed AI infrastructure environments that shorten the path from experimentation to production.