NetApp Novus targets AI metadata bottlenecks at scale

NetApp's Novus storage architecture delivers over 100TB/s aggregate throughput and separates metadata management from stored data, allowing each to scale independently. Its Data Director gives applications a unified view across multiple ONTAP clusters. The AI Data Engine discovers, classifies, and vectorizes enterprise data, aiming to cut preparation timelines that typically run six to nine months. NetApp's planned acquisition of PEAK:AIO is expected to add parallel file system technology based on the pNFS protocol, which is already supported in Linux kernels released after 2018, letting customers extend existing infrastructure without replacing it.