Amazon FSx for Lustre

Accelerate AI, ML, and HPC workloads with the fastest storage performance for GPU instances in the cloud, with up to 1200 Gbps per-client throughput with Elastic Fabric Adapter and NVIDIA GPUDirect Storage.

Starting at less than $0.005 per GB-month, the FSx for Lustre Intelligent-Tiering storage class helps you lower costs by only charging for the data you store, with automatic tiering between Frequent Access, Infrequent Access, and Archive tiers.

The only fully elastic Lustre file system automatically scales up or down based on your needs and only charges you for the data you store. You don’t need to worry about running out of capacity, updating hardware, managing file servers, configuring software, or tuning performance. Amazon FSx for Lustre is fully managed and automates these time-consuming administrative tasks.

Access and process Amazon Simple Storage Service (Amazon S3) data from a high-performance file system by linking your file systems to S3 buckets. Amazon FSx for Lustre also natively integrates with Amazon SageMaker HyperPod to provide fast storage for machine learning (ML) workloads and integrates with AWS services including Amazon EKS, AWS Backup, AWS ParallelCluster, Amazon CloudWatch, and IAM for comprehensive orchestration, monitoring, and security.

Accelerate ML data loading, model checkpointing, inferencing operations, and key-value (KV) caching with the fastest storage for GPU instances in the cloud.

Deliver terabytes per second of throughput and millions of IOPS to power high-performance workloads.

Accelerate quantitative analysis and risk modeling by processing massive financial datasets at high speeds.

Accelerate genomic analysis and cryo-EM analysis with high-performance storage designed for HCLS workflows.

Power precise weather predictions with high-throughput storage for processing meteorological models.