Storage

TIR Storage gives you three purpose-built options for AI/ML workflows. Use Datasets for object and disk-backed data management, SFS for shared concurrent access, and PFS for high-throughput parallel workloads.

DatasetsShared File-System (SFS)Parallel File-System (PFS)

Quick Start

Datasets

Manage data storage for AI/ML workflows with EOS and disk-backed options.

↗

Shared File-System (SFS)

Use centralized shared storage for concurrent multi-node access.

↗

Parallel File-System (PFS)

Use high-performance parallel storage for throughput-intensive workloads.

↗