Workload Placement helps you compare pricing options across cloud vendors.
- Specify a workload and we'll identify candidate instances with memory and cpu capacity
- We run benchmarking tasks across all major public cloud vendors to find machines with comparable computing power regardless of the underlying hardware, including many of the most common "on-prem" machines
- We also check for any unused RIs and reflect any custom pricing agreements you have calibrated within Cloudability to make sure you're paying the optimal price for a particular workload.
- On-Demand Price - The list price for a particular instance
- Reserved Price - The price after taking advantage of a commitment. In AWS and Azure, these are called reserved instances while in GCP the terminology is a "Committed-Use" discount. You can specify particular reservation options when you define your workload.
- Sustained-Use Price - GCP offers incremental discounts for increase usage of a particular instance type within a given month. You can set the estimated Usage Level when you define your workload. As AWS & Azure don't offer a similar concept, so this will reflect a sustained use discount of 0%.
- Spot Price - Azure, AWS and GCP offer highly discounted pricing for interruptible workloads. Note that Spot isn't available for all instance types in all regions.
Note that all prices reflect any custom pricing agreements you have calibrated within Cloudability.
Why am I seeing N/A in the pricing columns?
Each vendor typically offers a subset of pricing options in each region. We poll this information in real-time to give you the most up-to-date list of available discounts, but if you see "N/A" that means the pricing option you've selected is unavailable for a particular instance type and region.
Do you support Azure Hybrid Use Benefit?
We do! This benefit allows Azure customers to use on-premises Windows Server licenses in the cloud, so the underlying cost is equivalent to using a linux machine.
Why don't your memory specs match the AWS documentation?
Azure and GCP report instance memory in GB, while AWS reports memory in GiB. We report memory in GB here, based on an underlying passion about the one truly correct way to represent memory.
How can I use this to assist in a migration to the cloud?
We've included benchmarking statistics on many common on-premises machines. Simply select the hardware that most closely represents the machines you'd like to scope and we'll take care of the rest.