AI Infrastructure
AI data preprocessing: Where should it happen?
AI data preprocessing should happen where each transformation can run efficiently, within the data’s access and location constraints. A useful starting point ...
Scality Blog
Practical insight on exabyte-scale object storage, AI data infrastructure, and cyber resilience — from the team at Scality.
Object storage
Object storage sizing: Raw vs. usable capacity
Object storage sizing starts with the amount of data you need to retain, then accounts for the space required to protect it and keep the system operating. Raw ...
AI Infrastructure
GPU starvation: How to identify a storage bottleneck
GPU starvation occurs when an accelerator waits for the data or work it needs to continue processing. Storage can cause that wait, but so can image decoding, ...
Data sovereignty
Data sovereignty requirements: What to put in an RFP
Data sovereignty requirements in a request for proposal (RFP) should define where data may reside, who may access it, which legal entities are involved and how ...
Object storage
S3 migration: How to move object storage safely
S3 migration is the process of moving objects between storage systems that expose an S3 API. A safe migration preserves the data and the behavior applications ...
Cyber resilience
Backup deletion protection: Who can erase your copies?
Backup deletion protection prevents someone from removing recovery data before its required retention period ends. It depends on the permissions assigned to ...
Cyber resilience
Backup credential security: Separating access paths
Backup credential security means controlling the passwords, access keys and identities used to create, manage and restore backups. Its purpose is to prevent a ...
AI Infrastructure
AI storage sizing: How much capacity do you need?
AI storage sizing starts with everything your workloads need to retain: source datasets, prepared training data, saved dataset versions, checkpoints and model ...
AI Infrastructure
What storage architecture is best for RAG?
Retrieval-augmented generation (RAG) has become the default way to put an organization’s own documents in front of a large language model without retraining ...
AI Infrastructure
How should you store large datasets for AI training?
A training cluster is only as productive as the storage behind it. When accelerators sit idle because the data loader cannot read samples fast enough, or a ...