View a markdown version of this page

Select and prioritize workloads to move to AWS - Semiconductor Design on AWS

This whitepaper is for historical reference only. Some content might be outdated and some links might not be available.

Select and prioritize workloads to move to AWS

You will need to select and prioritize the workloads that you want to migrate to AWS. You may want to migrate a low-risk workflow first to use as a POC. After a successful POC, workloads that are time-sensitive or are constrained by the lack of adequate, on-premises compute resources are good candidates. Frequent data synchronization can be challenging and costly, so AWS suggests starting with workloads that have minimal runtime dependencies on on-premises data.

You may want to prioritize migrating workloads that use on-premises resources that are required for other purposes. AWS has customers that move both front-end and back-end workloads. Some customers run sign-off on AWS because it is critical to their schedule, and they lack the on-premises infrastructure needed to satisfy the resource requirements of these workloads. You might require only one workload to move to AWS, such as a bursty scale-out workload like IP characterization.

Another consideration is the data shared between workloads. You may benefit from keeping workloads that share data in the same location, either on-premises or in AWS.