NetApp launches StorageGRID 12.1, designed to support AI workloads and modern applications at scale. The latest version offers, among other things, up to twelve terabytes per second throughput.
NetApp launches StorageGRID 12.1, an update designed to support AI workloads and modern applications at scale. The new capabilities aim to improve how data is accessed, processed, and managed in distributed environments, supporting AI data pipelines, data lakes, and modern object-based applications.
“As organizations strive to turn rapidly growing and distributed volumes of unstructured data into insight and action, they need infrastructure that makes data intelligent, accessible, and AI-ready,” said Sandeep Singh, Senior Vice President and General Manager, Platform at NetApp. According to Forrester, an American market research firm, the platform is suitable for enterprises with complex hybrid and regulated data environments.
Federated global namespace
One of the key innovations in StorageGRID 12.1 is the federated global namespace. This allows customers to manage multiple StorageGRID systems worldwide without redesigning their applications or workflows. This functionality makes it possible to manage up to ten exabytes of data within a single namespace, which is crucial for organizations with dispersed data infrastructures.
The global namespace simplifies the management of distributed object storage and supports companies in centrally managing their data, regardless of physical location. This is essential for AI workloads that bring together large amounts of data from various sources for analysis and processing.
12 terabytes per second
According to NetApp, StorageGRID 12.1 delivers up to 400 percent more throughput than the previous version, depending on workload and object size. The maximum throughput is now twelve terabytes per second, enabling organizations to significantly accelerate their AI factories and data-intensive processes.
Furthermore, this version introduces batch operations on billions of objects and new features for AI agents. This allows changes in object storage to be efficiently tracked and enables the construction of comprehensive AI data pipelines without additional complexity.