S3 Standard-IA vs Intelligent-Tiering: Infrequent Access Done Right
Both classes keep millisecond retrieval while lowering the per-GB storage price. The difference is predictability: Standard-IA is cheaper storage with a retrieval fee, always; Intelligent-Tiering charges a small monitoring fee but automatically moves objects between tiers as access patterns change.
Standard-IA — for known rare access
Use Standard-IA when you know objects are accessed infrequently — backups you might restore, disaster-recovery copies, documents retrieved a few times a year. Storage price is roughly half of Standard; each retrieval carries a per-GB fee. Minimum storage duration (30 days) and minimum object size (128 KB) apply.
Intelligent-Tiering — for unknown or changing access
Intelligent-Tiering monitors access and moves objects between a frequent-access tier (Standard-priced), an infrequent tier (~Standard-IA-priced), and optional archive tiers — with no retrieval fees. Pay a small per-object monitoring charge instead. It's the answer for 'unknown or changing access pattern' scenarios.
Exam shortcuts
- 'Known infrequent access' → Standard-IA
- 'Unknown, changing, or unpredictable access' → Intelligent-Tiering
- 'Can tolerate hours-long retrieval' → Glacier Flexible Retrieval or Deep Archive (cheaper still)
- 'Must stay millisecond-latency' → eliminates the Glacier archive tiers