Data storage pioneer NetApp Inc. is building on its vision of an “intelligent data infrastructure” platform that’s fit for the artificial intelligence era with the launch of its new AI Data Engine.

Announced at the company’s annual user conference NetApp Insight 2025, it’s a comprehensive service that aims to make AI data pipelines simpler, easier to set up, more affordable and secure.

The NetApp AI Data Engine is designed to be used with a new disaggregated, all-flash storage system that’s dedicated to AI workloads. It’s called NetApp AFX, and the company wants it to become the foundation for “AI factories,” or essentially, the massive clusters of specialized computing infrastructure that powers AI models.

The company is also taking major steps towards securing those AI pipelines with new data breach detection and isolated recovery environments that can identify attempted breaches and prevent intrusions such as malware.

A foundation for AI data

NetApp said the new AI Data Engine is a secure and unified extension of the NetApp ONTAP operating system, which is designed for managing data storage across multiple environments, including the cloud, virtualized and on-premises settings. It’s based on Nvidia Corp.’s AI Data Platform reference design to help businesses simplify and secure their AI data pipelines and manage them through a single, unified control plane.

With the AI Data Engine, NetApp is trying to build the foundation of all AI data pipelines, taking care of both data ingestion and preparation, so it can serve any generative AI application or model, while providing a complete overview of customer’s entire data estates. It supports extensive search and data curation capabilities and features automatic data change detection and data synchronization to eliminate redundant copies and ensure that the information sent through pipelines is always up to date.

Because it’s based on Nvidia’s AI Platform reference design, it supports Nvidia-accelerated computing and specialized software such as NIM microservices, allowing it to search and retrieve vectorized information. Other capabilities include “advanced compression, fast semantic discovery and secure, policy-driven workflows,” the company said. It’s designed to run on top of the new AFX appliance, and will also support Nvidia’s RTX Pro servers featuring Blackwell Server Edition graphics processing units.

NetApp Chief Product Officer Syam Nair said customers will be able to use the AI Data Engine to connect their entire data estates across multicloud environments and build a unified data foundation for AI. “Enterprises can then dramatically accelerate their AI data pipelines by collapsing multiple data preparation and management steps into the integrated NetApp AI Data Engine,” he added.

The announcement is notable for being one of the first concrete steps taken by the company to try and fulfill its vision for AI, which is focused on helping organizations to become more data intelligent. In July, NetApp Chief Executive George Kurian (pictured) appeared on theCUBE, SiliconANGLE Media’s mobile livestreaming talk show, where he talked about how the company is trying to leverage its ability to unify siloed data across clouds and on-premises environments while preserving access controls, privacy rules and compliance mandates in order to support sensitive AI workloads:

Steve McDowell of NAND Research Inc. told SiliconANGLE that the AI Data Engine is based on an architecture that leverages NetApp’s core storage capabilities and fits them into a disaggregated solution that can support a broad range of AI and analytics workloads. “NetApp has talked about its NetApp Data Platform vision for AI for a long time, and today’s announcements are really the first instantiation of that vision,” he said.

The NetApp AFX appliance was built specifically for AI data workloads, decoupling performance and capacity through a disaggregated ONTAP. It’s certified to work with the Nvidia DGX SuperPOD supercomputing system for AI, offering all of NetApp’s robust data management and built-in cyber resilience capabilities, secure multi-tenancy, independent scaling of capacity and performance and compatibility with cloud and on-premises environments, the company said.

In addition, customers can also attach optional DX50 data control nodes to power a global metadata engine for a real-time catalog of enterprise data and leverage Nvidia accelerated computing. “The DX50 gives NetApp the flexibility to deploy new software features that align with enterprise AI needs,” McDowell said.

The analyst believes NetApp AFX can be a major differentiator for the company because it eliminates the headaches associated with trying to choose the best hardware for AI data. He explained that it’s able to handle some of the most complex data processing tasks, including metadata management and, of course, running the AI Data Engine.

“AFX is a compelling piece of engineering that makes it relatively easy to build a disaggregated data pipeline to meet most needs,” he said. “Where its competitors deliver a software-only solution that forces customers to figure out the hardware piece, NetApp provides packaged hardware components that just need to be connected together.”

Preventing data theft

To complement its new AI data foundation, NetApp announced a host of new cyber defense capabilities, with the most significant being its rebranded NetApp Ransomware Resilience service, which gains the ability to prevent data from being stolen.

Formerly known as the Ransomware Protection Service, NetApp Ransomware Resilience helps customers to identify and recover from ransomware attacks faster and more easily, while minimizing the damage they cause. The best thing about it is that it’s fully automated, supports both file and block storage, and can be controlled via a single control plane, without requiring deep security expertise or training, the company said.

Besides detecting ransomware, the service now comes with an AI-powered data breach detection capability that’s able to identify anomalous user and file system behavior that might indicate someone is trying to steal a company’s data. When it spots suspicious activity, it’ll issue an immediate alert to the customer through their integrated security information and event management tools, providing them with the necessary forensic data to take swift and decisive action. In this way, it can block the unauthorized movement of sensitive data almost as soon as it happens, preventing serious damage.

McDowell said NetApp’s ransomware detection capabilities were already notable, because it’s one of the only companies that tries to do this directly in the storage layer. “Nearly every other competitor, with IBM being the lone exception, detects malware in snapshots, not real time,” he said.

The analyst explained that NetApp is now building on top of this ability in order to prevent data being stolen. It’s a significant new capability that will go a long way towards reassuring enterprises, he believes, because after ransomware, stolen data is one of the biggest threats they face. “NetApp’s exfiltration detection is the first time we’ve seen this sort of ability integrated into a storage platform,” McDowell noted. “It’s a powerful feature that extends NetApp’s competitive moat in this area. It’s the only storage provider doing this.”

Another new feature in the Ransomware Protection Service is isolated recovery environments, which make it possible to cordon off individual workloads that may have come under ransomware attack. In these environments, NetApp’s AI scanning tools can then identify any maliciously impacted data and figure out the point at which it was modified. It’ll then guide customers step-by-step on what they need to do to restore that data safely and prevent reinfection.

Photo: SiliconANGLE