NetApp launches Novus to unclog enterprise AI data pipelines

NetApp has unveiled Novus, a new architecture designed to eliminate metadata bottlenecks in enterprise AI data infrastructure. The move targets a growing pain point: GPU clusters are scaling rapidly, but fragmented data and manual governance policies often stall AI workloads in pilot phases. Novus aims to unify storage and data management, making enterprise datasets accessible for models and agents without costly re-architecting.

The launch follows three years of positioning by NetApp, which has quietly repositioned its FlexPod joint venture with Cisco as a turnkey AI stack and framed sovereign storage as a backbone for European AI projects. Earlier StartupReader coverage noted that enterprises are increasingly prioritizing data infrastructure over model performance—a trend Novus appears built to exploit.

Novus’ pitch centers on automation: instead of manually tagging and governing datasets, the system uses metadata to dynamically organize and surface relevant data for AI workloads. This addresses a key tension in enterprise AI adoption: companies often lack the tools to operationalize their data at scale, even as GPU investments balloon.

NetApp’s bet is that AI’s next phase will be won by infrastructure players who can bridge the gap between raw compute power and usable data. If Novus delivers, it could accelerate AI deployments beyond proof-of-concept purgatory. Observers will watch whether the architecture gains traction—or remains another storage vendor’s AI gambit.

Sources: siliconangle.com