Reviewing the key considerations for optimal edge deployments
As we saw in the previous two chapters, there are several key factors that need to be taken into account when designing an appropriate architecture for training as well as deploying ML models at scale. In both these chapters, we also saw how Amazon SageMaker can be used to implement an effective ephemeral infrastructure for executing these tasks. Hence, in a later part of this chapter, we will also review how SageMaker can be used to deploy ML models to the edge at scale. Nonetheless, before we can dive into edge deployments with SageMaker, it is important to review some of the key factors that influence the successful deployment of an ML model at the edge:
- Efficiency
- Performance
- Reliability
- Security
While not all the mentioned factors may influence how an edge architecture is designed and may not be vital to the ML use case, it is important to at least consider them. So, let’s start by...
