|AWS (Taken from publicly available information from docs.aws.amazon.com)
- General Purpose
- Compute Optimised (c4, c5, c5a, c5ad, c5d, c5n, c6a, c6g, c6gd, c6gn, c6i, c6id, c7g). Each of them having anywhere between 5 and 10 variants. And AWS keeps modifying these offerings as they see the need
- Storage Optimised (d2, d3, d3en). Each of them having anywhere between 4 and 7 variants
- Accelerated Computing (DL1, F1, G3, G4ad, G4dn, G5, G5g). Each of them having 1 and 8 variants)
- Memory Optimised (R5, R5a, R5b, R5n, R6a, R6g, R6gd)
- HPC Instances
- As CSPs continue to optimise the offerings, the variants continue to explode.
- How to choose what's the best
- How often to do this exercise
- Given that majority of the customers have Hybrid cloud deployments, add AWS variants to what Azure has to offer, GCP has to offer, IBM Cloud has to offer and also, Alibaba, vmware.
- How feasible it is for the developer to take action, timely? And what's the measurement?