NetApp has launched Novus, a storage architecture designed to keep data flowing to large artificial intelligence GPU clusters and address storage bottlenecks that can reduce GPU utilisation to below 30%.
NetApp said Novus separates file metadata from data movement, allowing storage capacity and bandwidth to scale independently as AI infrastructure expands.
The initial release combines Novus Data Director running on Supermicro infrastructure with ONTAP services on NetApp AFF A90 systems.
NetApp said Novus is available to order and is designed to support high utilisation across clusters containing hundreds of thousands of GPUs.
Omdia said its testing showed near-linear performance gains as additional ONTAP clusters were added. Its modelling also projects sequential read throughput of 100 terabytes per second, rather than reporting that figure from a full-scale deployment.
“AI factories struggle and GPU economics collapse when data can’t keep up,” said NetApp chief product officer Syam Nair.
Separately, NetApp introduced AI-assisted operations in NetApp Console for organisations managing data across their own infrastructure and public clouds.
The company also launched Keystone Sovereign for qualified customers in the European Economic Area and said its integration with Nutanix Cloud Infrastructure for virtualised workloads is scheduled for general availability later this year.
