NetApp Launches Fastest Storage System, NetApp Novus, to Enhance AI Infrastructure Performance

NetApp® (NASDAQ: NTAP) officially launched the world's fastest storage system, NetApp Novus™, on October 6, 2026, designed for AI infrastructure providers operating large-scale GPU environments. This next-generation architecture supports zettabyte (ZB)-scale file systems, aimed at ensuring that "AI factories" can continuously supply data to millions of GPUs, thereby improving investment returns, GPU utilization, and profitability.

Addressing AI Factory Storage Bottlenecks

As emerging cloud and GPU-as-a-Service providers build massive AI factories, traditional storage architectures can no longer meet their needs, leading to declining GPU utilization and rising infrastructure costs. According to NetApp Chief Product Officer Syam Nair, when data supply cannot keep pace with compute demand, AI factories run into trouble, resulting in reduced GPU return on investment.

NetApp Novus delivers unlimited horizontal scaling of up to 100 TB/s through powerful performance, providing AI factories with the data support needed to unleash the performance potential of millions of GPUs. This all-new architecture not only meets current deployment requirements but also offers an evolution path to software-defined deployments, all while maintaining a single namespace, operational model, and data foundation.

Key Benefits and Architectural Features

Customers adopting the Novus architecture gain the following benefits:

• Ensure Efficient Operation of Every GPU: Novus maximizes GPU utilization by separating metadata services from the data path to reduce storage-related bottlenecks, continuously delivering high-performance data.
• Fragmentation-Free Scaling: Under a single namespace, Novus supports scaling to billions of files and massive AI datasets, allowing metadata, bandwidth, and capacity to scale independently, avoiding storage silos or increased operational complexity.
• Improve AI Infrastructure Efficiency: Novus enables operations personnel to scale resources for specific workloads, reducing power consumption, footprint, and infrastructure costs while preserving the flexibility, security, and multi-tenancy required for shared AI environments.

According to Mike Leone, Vice President and Principal Analyst at Moor Insights & Strategy, NetApp Novus integrates disaggregated metadata management with high-performance data services, providing a more scalable foundation for emerging cloud service and GPU-as-a-Service providers building AI factory operational scenarios.

NetApp Novus is now available for order, aimed at meeting the needs of multi-tenant, cloud-scale AI factories and providing strong support for future AI innovation.