Amazon File Cache
High-speed data cache
A high-speed cache (based on Lustre) on AWS that links to data in S3 or NFS (including on-prem) and loads data automatically when it is read, so workloads can process it quickly.
- Cost$$$ High
- Difficulty●●●●●
- Costs money when idle?Yes
- Managed by AWS
Think of it asLike a desk right next to the warehouse: bring the documents you need to the desk to work on quickly, then put them back.
When to use it
✓ Use it when
Burst workloads (rendering, ML, HPC) that need fast reads of data scattered across on-prem and S3.
✗ Avoid it when
Data is already in S3 and you just need a fast file system → FSx for Lustre.
Common uses
- Burst rendering in the cloud with on-prem data
- ML training on data spread across many locations
- Temporary genomics processing
What does it cost?
- Pricing model: Charged per hour by provisioned cache capacity; data transfer from the source is charged separately.
