Nebul Adopts NetApp Novus Storage Architecture for Sovereign AI Cloud
European AI cloud provider Nebul has become the first neocloud to adopt NetApp Novus storage, a high-performance architecture designed to support large-scale artificial intelligence infrastructure. Announced on November 10, the adoption brings NetApp's new storage technology into Nebul's plans for its sovereign AI cloud platform.
NetApp says Novus is architected for up to 100 terabytes per second (TB/s) of aggregate throughput. The system is designed to support AI factories operating with millions of GPUs, where moving data quickly and keeping computing resources supplied with information are major infrastructure challenges.
NetApp Novus storage targets large AI environments
NetApp introduced Novus in September 2026. Its architecture is designed around the changing requirements of neocloud providers and GPU-as-a-Service operators, which need to support multiple customers and demanding AI workloads.
A central feature is its disaggregated design. It allows performance, storage capacity and concurrency to scale independently under a single namespace. This approach is intended to help providers expand infrastructure as customer demand grows without requiring every part of the storage environment to scale in the same way.
The 100 TB/s figure represents the architecture's designed aggregate throughput capability. It should not be interpreted as a measured result from Nebul's infrastructure or the performance available to an individual customer.
Nebul plans to integrate Novus into its sovereign AI cloud
Nebul is developing what it calls the Intelligent Cloud, a European sovereign platform combining accelerated computing infrastructure, enterprise data services and production AI capabilities. Its services are intended to help organisations develop and operate AI applications while retaining control over their data and infrastructure.
The company expects NetApp Novus to support its multi-tenant AI environment, where different customers share infrastructure. The architecture is designed to help keep GPUs productive by allowing storage performance and capacity to expand independently as workloads increase.
Novus is built on NetApp ONTAP software. This could also help organisations already using the NetApp platform extend their existing data environments into Nebul's cloud while maintaining consistent security, data protection and governance policies. NetApp says the approach is intended to preserve existing data jurisdiction and access controls.
Why high-speed storage matters for AI infrastructure
AI infrastructure depends on more than processors. GPUs need continuous access to data during training and inference workloads. If storage systems cannot deliver information quickly enough, expensive computing resources may remain underused while waiting for data.
This makes storage architecture an important part of the wider AI infrastructure market. As neocloud providers expand GPU capacity, they also need systems that can handle higher throughput, multiple concurrent workloads and enterprise data requirements.
For Nebul, adopting NetApp Novus is part of its effort to build a sovereign cloud platform for organisations with demanding performance, security and compliance needs. The two companies are also exploring further opportunities to support enterprise adoption of sovereign AI services.
The announcement does not disclose Nebul's deployment schedule, the amount of storage capacity involved or measured performance figures from its platform. However, the adoption highlights how storage providers and AI cloud operators are working together to address the infrastructure demands of large-scale AI operations.
