Enabling Data using Nutanix Unified Storage

In their presentation at AI Field Day 7, Nutanix, represented by Distinguished Engineer Manoj Naik and SVP Vishal Sinha, laid out their vision for supporting AI workflows using Nutanix Unified Storage. They began by walking through a typical AI pipeline—from data ingestion and cleansing to model fine-tuning, inferencing, and archiving—emphasizing the challenges presented by data fragmentation across edge, cloud, and various formats. The Nutanix platform addresses these challenges by helping platform teams build and manage AI-ready data pipelines, ensuring clean, high-performance data is readily available for training large language models (LLMs) and inferencing operations.

The Nutanix solution includes a wide array of components that form a full-stack enterprise AI platform. Key pieces include Nutanix Cloud Infrastructure (NCI), Nutanix Unified Storage, the Nutanix Database Service (NDB), Nutanix Kubernetes Engine, and management layers such as Nutanix Central and Cloud Manager. Particularly noteworthy is their database management layer, which supports vector databases like PGVector and Milvus, enabling customers to manage LLM applications efficiently. The storage capabilities are tailored for each phase of the AI lifecycle—high-performance for training, rich metadata tiering for archiving, and versioning support via object storage and snapshotting.

The platform is designed to operate seamlessly across hybrid multicloud environments, offering data security, cyber resilience, and robust data mobility through features like global namespaces, immutable snapshots, and tiering based on metadata sensitivity. Nutanix Unified Storage supports multiple protocols (NFS, SMB, S3, iSCSI) and enables app and data colocation to optimize performance. The platform also supports cascading disaster recovery setups—including metro, asynchronous, and near-synchronous replication—to meet global compliance standards. With use cases ranging from edge AI inferencing to deep archival storage, Nutanix’s mature and versatile platform is built to handle the full lifecycle of data-intensive, AI-powered applications.