AI Cloud · Storage
High-throughput storage
Parallel filesystems and object storage tuned for AI workloads — hundreds of GB/s sustained throughput.
Shared Filesystem
High-speed NVMe shared filesystem — the default storage for our GPU compute, Managed Slurm and Kubernetes workloads.
Explore Shared Filesystem
Object Storage
S3-compatible object storage for model artefacts, dataset lakes and MLOps pipelines across our GPU platform.
Explore Object Storage
Storage sizing
Talk to our storage engineers about datasets from terabytes to exabytes.