NetApp Introduces Novus Architecture to Scale Production AI Workloads
CEO George Kurian outlines a decoupled metadata model designed to handle zettabyte-scale storage and high-throughput data pipelines.
NetApp Inc. introduced a new data architecture called Novus designed to support enterprise artificial intelligence workloads transitioning from experimental pilots into production environments.
Speaking during a broadcast on SiliconANGLE Media's theCUBE at NetApp INSIGHT, NetApp Chief Executive Officer George Kurian stated that 93% of organizations pursuing production AI cite data as their primary hurdle. Kurian argued that scaling AI across enterprise workflows introduces infrastructure requirements around data governance, security, and administrative control, as reported by SiliconANGLE .
To address these operational constraints, the Novus architecture pairs NetApp's established data layer with an independently scalable metadata layer. According to the company, the architecture is engineered to deliver transfer rates exceeding 100 terabytes per second and support zettabyte-scale storage within a single unified namespace.
Kurian explained that decoupling metadata processing allows organizations to perform data discovery, cataloging, and compliance guardrails without creating performance bottlenecks in raw data storage pipelines. The setup is designed to let enterprises deploy small initial installations and expand them incrementally to support larger AI environments on a single management platform.
The shift highlights growing infrastructure demands as companies aim to extract contextual memory from internal data repositories while maintaining strict access controls over proprietary intellectual property.
Sources
Written by
The Company Wire
Inside the companies building what’s next. Reporting on startups, technology, funding and the people shaping them.
